patent_num
int64 3.93M
10.2M
| claim_num1
int64 1
519
| claim_num2
int64 2
520
| sentence1
stringlengths 40
15.9k
| sentence2
stringlengths 88
20k
| label
float64 0.5
0.99
|
|---|---|---|---|---|---|
8,433,709 | 1 | 15 | 1. A modular lexical system for searching or inputting Chinese-like characters and words, the system having a processor configured to perform operations comprising: receiving input from a user to link or unlink one or more of a plurality of lexical data sources, each lexical data source having an independent data structure; generating one or more data structures for storing data from one or more of the lexical data sources in a plurality of data storage sections; indicating a structure of the stored data in one of the data storage sections; integrating the stored data in each data storage section into hierarchical data structure; creating an aggregate collection of lexemes, said collection of lexemes comprising an aggregate of all search keys and corresponding data found in the lexical data sources, together with cross-references to the lexical data sources in which said keys are found; designating, in response to user input, a subset of lexical data to be used; creating an activated subset of lexical data, said activated subset comprising a subset of the lexeme collection corresponding to the lexical data tables designated for use by the user, wherein each record corresponds to a phonetic or phonological search key for which the retrieved values correspond to character or word objects having an orthographic realization and additional lexical data as provided via the originating lexicon; retrieving individual records of lexical data from said lexical data sources such that all said sources may contribute candidates for input to the input means when a search key has been provided; displaying said candidates for selection by the user during word search and text input; generating a summary table that includes a basic nature and characteristic of individual lexicons stored in the system; and displaying information contained in said summary table. | 1. A modular lexical system for searching or inputting Chinese-like characters and words, the system having a processor configured to perform operations comprising: receiving input from a user to link or unlink one or more of a plurality of lexical data sources, each lexical data source having an independent data structure; generating one or more data structures for storing data from one or more of the lexical data sources in a plurality of data storage sections; indicating a structure of the stored data in one of the data storage sections; integrating the stored data in each data storage section into hierarchical data structure; creating an aggregate collection of lexemes, said collection of lexemes comprising an aggregate of all search keys and corresponding data found in the lexical data sources, together with cross-references to the lexical data sources in which said keys are found; designating, in response to user input, a subset of lexical data to be used; creating an activated subset of lexical data, said activated subset comprising a subset of the lexeme collection corresponding to the lexical data tables designated for use by the user, wherein each record corresponds to a phonetic or phonological search key for which the retrieved values correspond to character or word objects having an orthographic realization and additional lexical data as provided via the originating lexicon; retrieving individual records of lexical data from said lexical data sources such that all said sources may contribute candidates for input to the input means when a search key has been provided; displaying said candidates for selection by the user during word search and text input; generating a summary table that includes a basic nature and characteristic of individual lexicons stored in the system; and displaying information contained in said summary table. 15. The system of claim 1 , wherein the input language is one employing a letter-based writing system instead of Chinese-type characters. | 0.934762 |
8,401,282 | 1 | 5 | 1. A method for training a multi-class classifier, comprising the steps of: selecting a query image from a set of active images based on a membership probability determined by the classifier, wherein the active images are unlabeled; selecting a sample image from a set of training image based on the membership probability of the query image, wherein the training images are labeled; displaying the query image and the sample images to a user on an output device; obtaining a response from the user with an input device, wherein the response is a yes-match or a no-match; adding the query image with the label of the sample image to the training set if the yes-match is obtained, and otherwise repeating the selecting, displaying, and obtaining steps until a predetermined number of no-match is reached to obtain the multi-class classifier; applying the multi-class classifier to a set of unlabeled images to obtain a set of detection results; determining membership probabilities of the set of detection results; associating the set of detecting results with the membership probabilities less than a predetermined threshold with the set of active images; and retraining the multi-class classifier to refine the set of detection results, wherein steps of the method are performed by a processor. | 1. A method for training a multi-class classifier, comprising the steps of: selecting a query image from a set of active images based on a membership probability determined by the classifier, wherein the active images are unlabeled; selecting a sample image from a set of training image based on the membership probability of the query image, wherein the training images are labeled; displaying the query image and the sample images to a user on an output device; obtaining a response from the user with an input device, wherein the response is a yes-match or a no-match; adding the query image with the label of the sample image to the training set if the yes-match is obtained, and otherwise repeating the selecting, displaying, and obtaining steps until a predetermined number of no-match is reached to obtain the multi-class classifier; applying the multi-class classifier to a set of unlabeled images to obtain a set of detection results; determining membership probabilities of the set of detection results; associating the set of detecting results with the membership probabilities less than a predetermined threshold with the set of active images; and retraining the multi-class classifier to refine the set of detection results, wherein steps of the method are performed by a processor. 5. The method of claim 1 , further comprising: assigning a new class label to the query image if the predefined number of no-match is reached. | 0.770968 |
7,921,138 | 5 | 7 | 5. A computer memory having computer-readable components for localization of data included in software programs, the computer-readable components, when executed by at least one processor, comprising: data elements defined by a software data schema; and an owned comment data element created by an associated owner, the owned comment data element comprising information about the localization of the data included in the software programs, only the associated owner having permission to manipulate the owned comment data element, wherein: the owned comment data element includes a data field for enabling and disabling a processing of the owned comment data element, the owned comment data element is included in other data elements, and the owned comment data element is included in a list of owned comment elements comprising at least one owned comment data element; and the data elements being used in a computer-implemented method, the method comprising: when the data field for enabling and disabling the processing of the owned comment data element indicates that the owned comment data element is enabled, merging the owned comment data element with at least one another owned comment data element having ownership different from ownership of the owned comment data element, the merging comprising: when an ownership conflict between the owned comment data element and the at least one another owned comment data element does not arise, forming a single owned comment data element representing the owned comment data elements; and issuing an error message when during the merging an owned comment data element of the at least two owned comment data elements that is not associated with an owner comprising a file or a tool is disabled. | 5. A computer memory having computer-readable components for localization of data included in software programs, the computer-readable components, when executed by at least one processor, comprising: data elements defined by a software data schema; and an owned comment data element created by an associated owner, the owned comment data element comprising information about the localization of the data included in the software programs, only the associated owner having permission to manipulate the owned comment data element, wherein: the owned comment data element includes a data field for enabling and disabling a processing of the owned comment data element, the owned comment data element is included in other data elements, and the owned comment data element is included in a list of owned comment elements comprising at least one owned comment data element; and the data elements being used in a computer-implemented method, the method comprising: when the data field for enabling and disabling the processing of the owned comment data element indicates that the owned comment data element is enabled, merging the owned comment data element with at least one another owned comment data element having ownership different from ownership of the owned comment data element, the merging comprising: when an ownership conflict between the owned comment data element and the at least one another owned comment data element does not arise, forming a single owned comment data element representing the owned comment data elements; and issuing an error message when during the merging an owned comment data element of the at least two owned comment data elements that is not associated with an owner comprising a file or a tool is disabled. 7. The computer memory of claim 5 , wherein the owned comment data element comprises instructions relating to the localization of data included in the software programs for at least one of a software localization tool and a human operator. | 0.677898 |
7,797,674 | 16 | 17 | 16. A computing device for associating multiple domains within a block diagram model, the computing device comprising: a processor for executing: a plurality of partitioning blocks for use in defining a first region and a second region of a block diagram model; a first mechanism for associating a first domain with the first region of the block diagram model, the first domain comprising one or more nodes having a first function representing parameters associated with the one or more nodes of the first domain, parameters specific to the first domain being passed to the first region of the block diagram model without being passed to the second region of the block diagram model using the first function; and a second mechanism for associating a second domain with the second region of the block diagram model, the second domain comprising one or more nodes having a second function representing parameters associated with the one or more nodes of the second domain, parameters specific to the second domain being passed to the second region of the block diagram model without being passed to the first region of the block diagram model using the second function. | 16. A computing device for associating multiple domains within a block diagram model, the computing device comprising: a processor for executing: a plurality of partitioning blocks for use in defining a first region and a second region of a block diagram model; a first mechanism for associating a first domain with the first region of the block diagram model, the first domain comprising one or more nodes having a first function representing parameters associated with the one or more nodes of the first domain, parameters specific to the first domain being passed to the first region of the block diagram model without being passed to the second region of the block diagram model using the first function; and a second mechanism for associating a second domain with the second region of the block diagram model, the second domain comprising one or more nodes having a second function representing parameters associated with the one or more nodes of the second domain, parameters specific to the second domain being passed to the second region of the block diagram model without being passed to the first region of the block diagram model using the second function. 17. The computing device of claim 16 , wherein the plurality of partitioning blocks further comprises a manual configuration mechanism, the manual configuration mechanism allowing association of the first domain and the second domain with the first region and the second region of the block diagram model. | 0.526398 |
9,990,582 | 13 | 14 | 13. A non-transitory, computer-readable storage medium embodying computer program code, the computer program code comprising computer executable instructions configured for: storing data from a plurality of data sources within a cognitive graph via a cognitive inference and learning system, the cognitive graph representing a data domain; associating a first set of the data of the data domain within the cognitive graph with a first cognitive graph vector of a plurality of cognitive graph vectors via the cognitive inference and learning system, the first cognitive graph vector extending away from a cognitive graph nexus in a first direction; associating a second set of the data within the cognitive graph with a second cognitive graph vector of the plurality of cognitive graph vectors via the cognitive inference and learning system, the first cognitive graph vector extending away from a cognitive graph nexus in a first direction; processing the data from the plurality of data sources to provide cognitive insights via the cognitive inference and learning system; and refining the cognitive insights based upon a limitation relating to one of the plurality of cognitive graph vectors via the cognitive inference and learning system, the limitation corresponding to a selected cognitive graph vector parameter. | 13. A non-transitory, computer-readable storage medium embodying computer program code, the computer program code comprising computer executable instructions configured for: storing data from a plurality of data sources within a cognitive graph via a cognitive inference and learning system, the cognitive graph representing a data domain; associating a first set of the data of the data domain within the cognitive graph with a first cognitive graph vector of a plurality of cognitive graph vectors via the cognitive inference and learning system, the first cognitive graph vector extending away from a cognitive graph nexus in a first direction; associating a second set of the data within the cognitive graph with a second cognitive graph vector of the plurality of cognitive graph vectors via the cognitive inference and learning system, the first cognitive graph vector extending away from a cognitive graph nexus in a first direction; processing the data from the plurality of data sources to provide cognitive insights via the cognitive inference and learning system; and refining the cognitive insights based upon a limitation relating to one of the plurality of cognitive graph vectors via the cognitive inference and learning system, the limitation corresponding to a selected cognitive graph vector parameter. 14. The non-transitory, computer-readable storage medium of claim 13 , wherein: the first cognitive graph vector comprises a plurality of first cognitive graph vector indices extending along the first cognitive graph vector away from a cognitive graph nexus; the second cognitive graph vector comprises a plurality of second cognitive graph vector indices extending along the second cognitive graph vector away from the cognitive graph nexus; the limitation comprises limiting the first set of data to data within a first certain index of the plurality of first cognitive graph vector indices; and the refining comprising limiting the second set of data to data within a second certain index of the second cognitive graph vector indices. | 0.5 |
8,375,067 | 1 | 2 | 1. A system comprising: a plurality of processors; a database comprising data related to a plurality of job seekers and/or a plurality of jobs; a matching module implemented on one of the plurality of processors in communication with the database, the matching module producing job matching results by matching one of the job seekers to a potential job through finding one or more common parameters between job seeker parameters associated with the one job seeker and one or more job parameters associated with the potential job; an individualized data collection module determining job seeker characteristics based on individualized information captured from a user interface instantiated on a job seeker's individual device, wherein the determined job seeker characteristics are incorporated into job seeker parameters utilized by the matching module; and a correlation module implemented on one of the plurality of processors in communication with the matching module, the correlation module determining a correlation between one of the common parameters and one or more alternative jobs from the plurality of jobs in part based on tracked behavior characteristics of prior interactions with the system for each job seeker, wherein the correlation module comprises a filtration module generating and storing a negative filtration indication associated with a job excluding the job in the job matching results and the correlated one or more alternative jobs responsive to a job seeker input negative indication related to the excluded job, and wherein the correlation module further comprises an affinity engine configured to determine affinity metrics between the potential job identified in the job matching results and other jobs interacted with by the plurality of job seekers and to determine alternative jobs to the job matching results based on the determined affinity metrics. | 1. A system comprising: a plurality of processors; a database comprising data related to a plurality of job seekers and/or a plurality of jobs; a matching module implemented on one of the plurality of processors in communication with the database, the matching module producing job matching results by matching one of the job seekers to a potential job through finding one or more common parameters between job seeker parameters associated with the one job seeker and one or more job parameters associated with the potential job; an individualized data collection module determining job seeker characteristics based on individualized information captured from a user interface instantiated on a job seeker's individual device, wherein the determined job seeker characteristics are incorporated into job seeker parameters utilized by the matching module; and a correlation module implemented on one of the plurality of processors in communication with the matching module, the correlation module determining a correlation between one of the common parameters and one or more alternative jobs from the plurality of jobs in part based on tracked behavior characteristics of prior interactions with the system for each job seeker, wherein the correlation module comprises a filtration module generating and storing a negative filtration indication associated with a job excluding the job in the job matching results and the correlated one or more alternative jobs responsive to a job seeker input negative indication related to the excluded job, and wherein the correlation module further comprises an affinity engine configured to determine affinity metrics between the potential job identified in the job matching results and other jobs interacted with by the plurality of job seekers and to determine alternative jobs to the job matching results based on the determined affinity metrics. 2. The system according to claim 1 wherein the matching module determines a matching score for each of the common parameters between the job seeker and the one or more jobs. | 0.524725 |
7,788,142 | 2 | 3 | 2. The computer readable media as recited in claim 1 , wherein the instructions perform a further step comprising examining prior purchasing histories of each of the multiple customers having the keyword in their company name data and additionally recommending to the potential customer one or more products commonly found in the prior purchasing histories of the multiple customers having the keyword in their company name data. | 2. The computer readable media as recited in claim 1 , wherein the instructions perform a further step comprising examining prior purchasing histories of each of the multiple customers having the keyword in their company name data and additionally recommending to the potential customer one or more products commonly found in the prior purchasing histories of the multiple customers having the keyword in their company name data. 3. The computer readable media as recited in claim 2 , wherein the instructions perform a further step comprising aggregating the item and the one or more products into a catalog customized for the potential customer. | 0.5 |
7,509,303 | 25 | 26 | 25. The computer-implemented process of claim 21 , wherein said collecting is performed at a frequency that is data source dependent. | 25. The computer-implemented process of claim 21 , wherein said collecting is performed at a frequency that is data source dependent. 26. The computer-implemented process of claim 25 , wherein said collecting is performed at a frequency based on a type of said data source. | 0.5 |
7,487,448 | 14 | 16 | 14. The system of claim 11 , wherein the grouping element has properties that can be expressed multiple ways. | 14. The system of claim 11 , wherein the grouping element has properties that can be expressed multiple ways. 16. The system of claim 14 , wherein one way in which a grouping element property can be expressed is via a resource dictionary reference. | 0.743494 |
9,785,658 | 13 | 18 | 13. A system comprising: a memory that stores instructions; and one or more processors configured by the instructions to perform operations comprising: accessing a first schema comprising a first plurality of business entities including a first business entity, the first business entity having a first name in the first schema; accessing a second schema comprising a second plurality of business entities including the first business entity, the first business entity having a second name in the second schema; generating from the first schema and the second schema, a merged schema comprising a third plurality of business entities, including a single instance of the first business entity; storing the merged schema in a database; extracting a first sequence of words from the first name of the first business entity; generating candidate phrases for the business entity from the first and second sequences of words; and ranking the candidate phrases for the first business entity; analyzing candidate sets of labels for the third plurality of business entities, each candidate set of labels including a label for each business entity of the third plurality of business entities, no two business entities having the same label in the candidate set of labels, the label for the first business entity being selected from the candidate phrases for the first business entity; and assigning labels to each business entity of the third plurality of business entities based on the analysis of the candidate sets of labels; and receiving data stored using the first schema; converting the received data to the merged schema; and causing a presentation of the converted data using the assigned labels. | 13. A system comprising: a memory that stores instructions; and one or more processors configured by the instructions to perform operations comprising: accessing a first schema comprising a first plurality of business entities including a first business entity, the first business entity having a first name in the first schema; accessing a second schema comprising a second plurality of business entities including the first business entity, the first business entity having a second name in the second schema; generating from the first schema and the second schema, a merged schema comprising a third plurality of business entities, including a single instance of the first business entity; storing the merged schema in a database; extracting a first sequence of words from the first name of the first business entity; generating candidate phrases for the business entity from the first and second sequences of words; and ranking the candidate phrases for the first business entity; analyzing candidate sets of labels for the third plurality of business entities, each candidate set of labels including a label for each business entity of the third plurality of business entities, no two business entities having the same label in the candidate set of labels, the label for the first business entity being selected from the candidate phrases for the first business entity; and assigning labels to each business entity of the third plurality of business entities based on the analysis of the candidate sets of labels; and receiving data stored using the first schema; converting the received data to the merged schema; and causing a presentation of the converted data using the assigned labels. 18. The system of claim 13 , wherein the ranking of the candidate phrases for the first business entity includes ranking the candidate phrases for the first business entity based on a length of each candidate phrase. | 0.788235 |
9,786,268 | 17 | 19 | 17. The apparatus of claim 16 , wherein the processor is further configured to: capture the related text and locate the related information on an associated video stored in the one or more apparatus and the social media website. | 17. The apparatus of claim 16 , wherein the processor is further configured to: capture the related text and locate the related information on an associated video stored in the one or more apparatus and the social media website. 19. The apparatus of claim 17 , wherein the processor is further configured to: determine whether the apparatus is a mobile phone or a computer system; and access a file system of the apparatus based on whether the apparatus is the mobile phone or the computer system. | 0.782114 |
8,954,845 | 1 | 7 | 1. An image processing device comprising: at least one processor and memory, cooperating to function as: an input unit configured to input document image data; a region division unit configured to divide the document image data into a plurality of regions according to attributes, the divided regions including a text region, a caption region and an object region which is accompanied by the caption region; a character recognition unit configured to obtain character information by executing a character recognition process for each character within each of the text region and the caption region divided by said region division unit; an anchor expression extraction unit configured to extract, from the character information in the caption region, an anchor expression which includes a predetermined character string identifying the object region; a text search unit configured to search for the anchor expression extracted by said anchor expression extraction unit from the character information in the text region; a link information generation unit configured to generate two-way link information associating an anchor expression peripheral region and an image peripheral region with each other, the anchor expression peripheral region being a region including the anchor expression found by said text search unit, and the image peripheral region being a region including the object region, wherein, in a case where said text search unit finds one anchor expression from the character information in the text region, said link information generation unit generates the two-way link information associating the image peripheral region with one anchor expression peripheral region which is a region including the one anchor expression found by said text search unit, wherein, in a case where said text search unit finds a plurality of anchor expressions from the character information in the text region, said link information generation unit generates the two-way link information associating the image peripheral region with a plurality of anchor expression peripheral regions which are regions including the plurality of the anchor expressions found by said text search unit; and a format conversion unit configured to generate electronic document data including document image data and the two-way link information, wherein, in the case where said text search unit finds the one anchor expression from the character information in the text region, the two-way link information includes a first link and a second link, the first link including information for displaying the associated one anchor expression peripheral region when a reader of the electronic document takes a predetermined action on the image peripheral region, and the second link including information for displaying the associated image peripheral region when a reader of the electronic document takes a predetermined action on the one anchor expression peripheral region, and wherein, in the case where said text search unit finds the plurality of anchor expressions from the character information in the text region, the two-way link information includes a third link and a fourth link, the third link including information for displaying information about the plurality of the anchor expression peripheral regions as a plurality of candidates of link destinations from the image peripheral region when a reader of the electronic document takes a predetermined action on the image peripheral region, and the fourth link including information for displaying the associated image peripheral region when a reader of the electronic document takes a predetermined action on any one of the plurality of the anchor expression peripheral regions. | 1. An image processing device comprising: at least one processor and memory, cooperating to function as: an input unit configured to input document image data; a region division unit configured to divide the document image data into a plurality of regions according to attributes, the divided regions including a text region, a caption region and an object region which is accompanied by the caption region; a character recognition unit configured to obtain character information by executing a character recognition process for each character within each of the text region and the caption region divided by said region division unit; an anchor expression extraction unit configured to extract, from the character information in the caption region, an anchor expression which includes a predetermined character string identifying the object region; a text search unit configured to search for the anchor expression extracted by said anchor expression extraction unit from the character information in the text region; a link information generation unit configured to generate two-way link information associating an anchor expression peripheral region and an image peripheral region with each other, the anchor expression peripheral region being a region including the anchor expression found by said text search unit, and the image peripheral region being a region including the object region, wherein, in a case where said text search unit finds one anchor expression from the character information in the text region, said link information generation unit generates the two-way link information associating the image peripheral region with one anchor expression peripheral region which is a region including the one anchor expression found by said text search unit, wherein, in a case where said text search unit finds a plurality of anchor expressions from the character information in the text region, said link information generation unit generates the two-way link information associating the image peripheral region with a plurality of anchor expression peripheral regions which are regions including the plurality of the anchor expressions found by said text search unit; and a format conversion unit configured to generate electronic document data including document image data and the two-way link information, wherein, in the case where said text search unit finds the one anchor expression from the character information in the text region, the two-way link information includes a first link and a second link, the first link including information for displaying the associated one anchor expression peripheral region when a reader of the electronic document takes a predetermined action on the image peripheral region, and the second link including information for displaying the associated image peripheral region when a reader of the electronic document takes a predetermined action on the one anchor expression peripheral region, and wherein, in the case where said text search unit finds the plurality of anchor expressions from the character information in the text region, the two-way link information includes a third link and a fourth link, the third link including information for displaying information about the plurality of the anchor expression peripheral regions as a plurality of candidates of link destinations from the image peripheral region when a reader of the electronic document takes a predetermined action on the image peripheral region, and the fourth link including information for displaying the associated image peripheral region when a reader of the electronic document takes a predetermined action on any one of the plurality of the anchor expression peripheral regions. 7. The image processing device of claim 1 , wherein the document image data included in the generated electronic document data is data that is obtained by executing at least one of vector conversion processing and image compress processing. | 0.737991 |
9,223,399 | 11 | 15 | 11. A method for translating gestures in a virtual world, comprising: receiving an input from a first user representing an input gesture to be made by a first avatar to a second avatar in the virtual world; translating, using a processor, the input gesture input, by the first user, based at least in part on one or more environmental, cultural or social factors to generate at least one translated gesture; outputting on a first display a depiction of the gesture input by the first user as being made by the first avatar to the second avatar; and outputting on a second display the translated gesture as being made by the first avatar to the second avatar, wherein the first display is separate from the second display. | 11. A method for translating gestures in a virtual world, comprising: receiving an input from a first user representing an input gesture to be made by a first avatar to a second avatar in the virtual world; translating, using a processor, the input gesture input, by the first user, based at least in part on one or more environmental, cultural or social factors to generate at least one translated gesture; outputting on a first display a depiction of the gesture input by the first user as being made by the first avatar to the second avatar; and outputting on a second display the translated gesture as being made by the first avatar to the second avatar, wherein the first display is separate from the second display. 15. The method of claim 11 , wherein translating the input gesture input by the first user to generate at least one translated gesture comprises translating the input gesture based on a virtual environment in which either the first and/or second avatar may be located. | 0.5 |
8,311,802 | 8 | 10 | 8. The method of claim 1 , wherein the second portion corresponds to a second language. | 8. The method of claim 1 , wherein the second portion corresponds to a second language. 10. The method of claim 8 , wherein the first language comprises a Roman language and the second language comprises a non-Roman language. | 0.5 |
8,935,303 | 1 | 9 | 1. A method of optimizing an output ranked list of recommended items given an input user, an input item list, and an input context, comprising: providing a multidimensional data set that comprises information of interactions from a plurality of users with a plurality of items and in a plurality of contexts; factorizing the multidimensional data set into a number of two-dimensional matrices, the number of two-dimensional matrices being equivalent to the number of dimensions that the multidimensional data set has; computing a mathematical recommendation model by optimizing an objective function over the two-dimensional matrices into which the multidimensional data set has been factorized, the recommendation model comprising a score value for each combination of user, item and context; and computing the output ranked list by applying the computed recommendation model to the input user, input item list and input context, wherein the recommendation model further comprises a ranked list of recommended items for each user and context, being each ranked list determined by sorting the scores of the plurality of items for each user and context; and wherein the objective function is a continuous function with infinite continuous derivatives that quantifies a relevance of the recommended items of each ranked list of the recommendation model, calculated over at least some of the plurality of users and over at least some of the plurality of contexts. | 1. A method of optimizing an output ranked list of recommended items given an input user, an input item list, and an input context, comprising: providing a multidimensional data set that comprises information of interactions from a plurality of users with a plurality of items and in a plurality of contexts; factorizing the multidimensional data set into a number of two-dimensional matrices, the number of two-dimensional matrices being equivalent to the number of dimensions that the multidimensional data set has; computing a mathematical recommendation model by optimizing an objective function over the two-dimensional matrices into which the multidimensional data set has been factorized, the recommendation model comprising a score value for each combination of user, item and context; and computing the output ranked list by applying the computed recommendation model to the input user, input item list and input context, wherein the recommendation model further comprises a ranked list of recommended items for each user and context, being each ranked list determined by sorting the scores of the plurality of items for each user and context; and wherein the objective function is a continuous function with infinite continuous derivatives that quantifies a relevance of the recommended items of each ranked list of the recommendation model, calculated over at least some of the plurality of users and over at least some of the plurality of contexts. 9. The method of claim 1 wherein the interactions whose information is comprised in the multidimensional data set are selected from a group of implicit feedbacks comprising: a click, a mouse movements, a purchase, an installation of an application, a browsing history, an usage history and a search pattern. | 0.5 |
9,928,227 | 7 | 11 | 7. A form filling system, comprising: a cache memory; and a processor coupled to the cache memory, wherein the processor is configured to: receive a message; analyze content of the message utilizing natural language processing (NLP) to locate an action command within the message; apply NLP to text associated with the action command to determine a context of the action command; select multiple candidate forms based on the multiple candidate forms being preconfigured to allow completion of the action command; apply a similarity algorithm to the text associated with the action command and one or more input fields of the multiple candidate forms to identify one or more matching elements between the text and the input fields of the forms; select a form, from among the multiple candidate forms, based on which of the multiple candidate forms has a greatest number of the matching elements between the text and the input fields of the form; and automatically pre-fill the input fields of the form with information associated with the matching elements in the message whose similarity exceeds a predetermined threshold without user interaction to minimize copy errors and reduce task execution time associated with filling in the form. | 7. A form filling system, comprising: a cache memory; and a processor coupled to the cache memory, wherein the processor is configured to: receive a message; analyze content of the message utilizing natural language processing (NLP) to locate an action command within the message; apply NLP to text associated with the action command to determine a context of the action command; select multiple candidate forms based on the multiple candidate forms being preconfigured to allow completion of the action command; apply a similarity algorithm to the text associated with the action command and one or more input fields of the multiple candidate forms to identify one or more matching elements between the text and the input fields of the forms; select a form, from among the multiple candidate forms, based on which of the multiple candidate forms has a greatest number of the matching elements between the text and the input fields of the form; and automatically pre-fill the input fields of the form with information associated with the matching elements in the message whose similarity exceeds a predetermined threshold without user interaction to minimize copy errors and reduce task execution time associated with filling in the form. 11. The form filling system of claim 7 , wherein the message is one of a text message, an email message, a message included in a word processing document, and a message generated by a web page. | 0.678333 |
9,342,626 | 1 | 11 | 1. A method, comprising: identifying a current query of a user, wherein the current query is a partial query entered by the user; identifying one or more past queries of the user, the past queries issued by the user prior to the current query; identifying one or more past entity collections related to one or more of the identified past queries, the past entity collections being a first set of entity collections, wherein each of the entity collections includes a grouping of entities that are members of the entity collection and that share one or more aspects in common; identifying one or more candidate query suggestions for the current query based at least in part on one or more characters of the current query; identifying, for a given candidate query suggestion of the candidate query suggestions, one or more current entity collections related to the given candidate query suggestion, the current entity collections being a second set of the entity collections; determining, for the given candidate query suggestion, the current entity collections that match the past entity collections; and ranking the given candidate query suggestion based on a comparison of the current entity collections that match the past entity collections to the current entity collections of a group of the current entity collections, the group including one or more of the current entity collections that do not match the past entity collections. | 1. A method, comprising: identifying a current query of a user, wherein the current query is a partial query entered by the user; identifying one or more past queries of the user, the past queries issued by the user prior to the current query; identifying one or more past entity collections related to one or more of the identified past queries, the past entity collections being a first set of entity collections, wherein each of the entity collections includes a grouping of entities that are members of the entity collection and that share one or more aspects in common; identifying one or more candidate query suggestions for the current query based at least in part on one or more characters of the current query; identifying, for a given candidate query suggestion of the candidate query suggestions, one or more current entity collections related to the given candidate query suggestion, the current entity collections being a second set of the entity collections; determining, for the given candidate query suggestion, the current entity collections that match the past entity collections; and ranking the given candidate query suggestion based on a comparison of the current entity collections that match the past entity collections to the current entity collections of a group of the current entity collections, the group including one or more of the current entity collections that do not match the past entity collections. 11. The method of claim 1 , wherein identifying the one or more current entity collections includes: determining, for each of the current entity collections, that a ranking of the entity collection for the current query satisfies a threshold. | 0.745263 |
10,114,612 | 1 | 11 | 1. A method comprising: constructing, via a processor device, a media interconnection graph which models how media are interconnected by connecting disparate categories of the media; normalizing the media interconnection graph, to yield a normalized graph; generating, via the processor device, a speech recognition model based on the normalized graph; and receiving audible speech via a microphone to yield received audible speech for searching media content; converting, via the processor device, the received audible speech to output a graph using the speech recognition model. | 1. A method comprising: constructing, via a processor device, a media interconnection graph which models how media are interconnected by connecting disparate categories of the media; normalizing the media interconnection graph, to yield a normalized graph; generating, via the processor device, a speech recognition model based on the normalized graph; and receiving audible speech via a microphone to yield received audible speech for searching media content; converting, via the processor device, the received audible speech to output a graph using the speech recognition model. 11. The method of claim 1 , wherein the speech recognition model is a hierarchical language model. | 0.812261 |
7,711,673 | 13 | 18 | 13. A computer-implemented method for handling an email message received through a communication network, said email message including a received document, said received document involving an encoding scheme, the method comprising: receiving a plurality of text document samples, said plurality of text document samples being encoded with different encoding schemes and selected for training purposes, said different encoding schemes pertaining to charset encoding for transmission over a network; and training, using said plurality of text document samples, to obtain a set of machine learning models, said training including generating fundamental units from said plurality of text document samples for charsets of said plurality of text document samples, extracting a subset of said fundamental units as feature lists, said extracting said subset of said fundamental units including filtering said fundamental units to obtain fundamental units that are more discriminatory in describing differences between said different encoding schemes, converting said feature lists into a set of feature vectors, and generating said set of machine learning models from said set of feature vectors using SIM (Similarity Algorithm), said feature vectors are grouped by charsets; applying said set of machine learning models against a set of received document feature vectors converted from said received document, said applying including analyzing said set of received document feature vectors using said set of machine learning models to compute similarity indicia between said set of received document feature vectors and said set of machine learning models associated with said different encoding schemes, said similarity indicia including at least a set of cross-angles between said set of target document feature vectors and said set of machine learning models, wherein a first encoding scheme associated with said set of machine learning models is designated as said encoding scheme if characteristics of said first encoding scheme as represented by said set of machine learning models are computed to be most similar, relative to other encoding schemes of said different encoding schemes, to said set of received document feature vectors; decoding said target document to obtain decoded content of said document based on at least said first encoding scheme; determining whether said email message is a spam message based on at least said decoded content of said document; and preventing said email message from reaching an email user if said email message is determined to be spam according to said determining. | 13. A computer-implemented method for handling an email message received through a communication network, said email message including a received document, said received document involving an encoding scheme, the method comprising: receiving a plurality of text document samples, said plurality of text document samples being encoded with different encoding schemes and selected for training purposes, said different encoding schemes pertaining to charset encoding for transmission over a network; and training, using said plurality of text document samples, to obtain a set of machine learning models, said training including generating fundamental units from said plurality of text document samples for charsets of said plurality of text document samples, extracting a subset of said fundamental units as feature lists, said extracting said subset of said fundamental units including filtering said fundamental units to obtain fundamental units that are more discriminatory in describing differences between said different encoding schemes, converting said feature lists into a set of feature vectors, and generating said set of machine learning models from said set of feature vectors using SIM (Similarity Algorithm), said feature vectors are grouped by charsets; applying said set of machine learning models against a set of received document feature vectors converted from said received document, said applying including analyzing said set of received document feature vectors using said set of machine learning models to compute similarity indicia between said set of received document feature vectors and said set of machine learning models associated with said different encoding schemes, said similarity indicia including at least a set of cross-angles between said set of target document feature vectors and said set of machine learning models, wherein a first encoding scheme associated with said set of machine learning models is designated as said encoding scheme if characteristics of said first encoding scheme as represented by said set of machine learning models are computed to be most similar, relative to other encoding schemes of said different encoding schemes, to said set of received document feature vectors; decoding said target document to obtain decoded content of said document based on at least said first encoding scheme; determining whether said email message is a spam message based on at least said decoded content of said document; and preventing said email message from reaching an email user if said email message is determined to be spam according to said determining. 18. The computer-implemented method of claim 13 further comprising displaying said decoded content if said email message is determined to be not spam according to said determining. | 0.65251 |
9,619,024 | 1 | 6 | 1. A virtual inputting device, comprising: a signal collection unit including a bioelectrical sensor for collecting bioelectrical signals and an acceleration sensor for collecting acceleration signals, the bioelectrical signals and the acceleration signals reflecting a user's gesture; a signal preprocessing unit for performing preprocessing for the bioelectrical signals and the acceleration signals collected by the signal collection unit; a signal segmentation unit for performing segmentation processing for the preprocessed bioelectrical signals and acceleration signals so as to obtain a plurality of gesture segments; a feature extracting unit for extracting feature values from the bioelectrical signals and the acceleration signals for respective gesture segments; a feature combination unit for combining feature values extracted from the feature extracting unit to form a combined feature vector; a gesture recognition unit for performing gesture recognition based on the combined feature vector; and a character mapping unit for obtaining characters corresponding to the recognized gesture according to a predetermined mapping relationship between characters and gestures, wherein the segmentation processing comprises: determining starting points and ending points for the preprocessed bioelectrical signals and the preprocessed acceleration signals respectively; and averaging the starting points so as to obtain a starting point of a gesture segment, and averaging the ending points so as to obtain an ending point of the gesture segment. | 1. A virtual inputting device, comprising: a signal collection unit including a bioelectrical sensor for collecting bioelectrical signals and an acceleration sensor for collecting acceleration signals, the bioelectrical signals and the acceleration signals reflecting a user's gesture; a signal preprocessing unit for performing preprocessing for the bioelectrical signals and the acceleration signals collected by the signal collection unit; a signal segmentation unit for performing segmentation processing for the preprocessed bioelectrical signals and acceleration signals so as to obtain a plurality of gesture segments; a feature extracting unit for extracting feature values from the bioelectrical signals and the acceleration signals for respective gesture segments; a feature combination unit for combining feature values extracted from the feature extracting unit to form a combined feature vector; a gesture recognition unit for performing gesture recognition based on the combined feature vector; and a character mapping unit for obtaining characters corresponding to the recognized gesture according to a predetermined mapping relationship between characters and gestures, wherein the segmentation processing comprises: determining starting points and ending points for the preprocessed bioelectrical signals and the preprocessed acceleration signals respectively; and averaging the starting points so as to obtain a starting point of a gesture segment, and averaging the ending points so as to obtain an ending point of the gesture segment. 6. The virtual input device of claim 1 , wherein the preprocessing comprises band-pass filtering for the bioelectrical signals, low pass filtering for the acceleration signals, and analog to digital conversion for the bioelectrical signals and the acceleration signals. | 0.752757 |
9,122,825 | 5 | 6 | 5. A computer-readable storage medium storing computer program instructions that, when executed on a processor, carry out actions including: analyzing the hierarchical design using a processor to determine dependency chains from leaf cells that do not reference any other cells in the hierarchical design upward to higher levels in the hierarchical design, wherein n-level cells, by definition, are limited to referencing leaf cells through n−1 level cells and not referencing n−level or n+level cells; detecting matching leaf cells, assigning matching names to the matching leaf cells, and propagating the matching names to higher level cells in the hierarchical design in place of the original, non-matching names of the matching leaf cells; applying the detecting, assigning and propagating steps to at least some first-level, second-level, third-level, and subsequent-level cells along the dependency chains; and evaluating functional similarity of at least some of the n-level cells above the leaf cells in the hierarchical design after propagating the matching names into the n-level cells along the dependency chains. | 5. A computer-readable storage medium storing computer program instructions that, when executed on a processor, carry out actions including: analyzing the hierarchical design using a processor to determine dependency chains from leaf cells that do not reference any other cells in the hierarchical design upward to higher levels in the hierarchical design, wherein n-level cells, by definition, are limited to referencing leaf cells through n−1 level cells and not referencing n−level or n+level cells; detecting matching leaf cells, assigning matching names to the matching leaf cells, and propagating the matching names to higher level cells in the hierarchical design in place of the original, non-matching names of the matching leaf cells; applying the detecting, assigning and propagating steps to at least some first-level, second-level, third-level, and subsequent-level cells along the dependency chains; and evaluating functional similarity of at least some of the n-level cells above the leaf cells in the hierarchical design after propagating the matching names into the n-level cells along the dependency chains. 6. The computer-readable storage medium of claim 5 , further including program instructions to carry out speculative matching of cells, including: detecting leaf cells that appear to be modified counterparts of one another and assigning matching names to the modified counterpart leaf cells to replace original, non-matching names of the modified counterpart leaf cells, before the propagating of the matching names; applying the detecting, assigning and propagating steps for modified counterparts to at least some first-level, second-level, third-level and subsequent-level cells along the dependency chains; and evaluating functional similarity of at least some of the n-level cells above the leaf cells in the hierarchical design after propagating the matching names into the higher-level cells along the dependency chains. | 0.5 |
8,792,142 | 1 | 3 | 1. A method of creating a structural document, the method comprising: receiving, by a host computing device in a cloud system, content information pertaining to one or more contents that are to be encased by the structural document; determining, by the host computing device, a shape of a structural document based on the received content information; determining, by the host computing device, a plurality of dimensions of the structural document based on the received content information; receiving, by the host computing device, information associated with one or more content items, wherein each content item represents a visual characteristic associated with the structural document; receiving, via a communications network from a user computing device that is remote from the host computing device, contact information to include on the structural document, wherein the contact information corresponds to a product information provider associated with the structural document; receiving, by the host computing device, an indication of a location on the structural document where the contact information is to be displayed; causing a graphical representation of the structural document to be displayed at a user computing device, wherein a shape of the three-dimensional graphical representation corresponds to the determined shape, wherein a plurality of dimensions of the graphical representation correspond to the determined plurality of dimensions, wherein the contact information is displayed on the graphical representation at the location, wherein the graphical representation comprises at least a portion of the received content items; receiving, from the user computing device by the host computing device, an indication that a user is finished creating the structural document; generating a print document comprising an encoded data mark; and providing the print document to one or more print-related devices. | 1. A method of creating a structural document, the method comprising: receiving, by a host computing device in a cloud system, content information pertaining to one or more contents that are to be encased by the structural document; determining, by the host computing device, a shape of a structural document based on the received content information; determining, by the host computing device, a plurality of dimensions of the structural document based on the received content information; receiving, by the host computing device, information associated with one or more content items, wherein each content item represents a visual characteristic associated with the structural document; receiving, via a communications network from a user computing device that is remote from the host computing device, contact information to include on the structural document, wherein the contact information corresponds to a product information provider associated with the structural document; receiving, by the host computing device, an indication of a location on the structural document where the contact information is to be displayed; causing a graphical representation of the structural document to be displayed at a user computing device, wherein a shape of the three-dimensional graphical representation corresponds to the determined shape, wherein a plurality of dimensions of the graphical representation correspond to the determined plurality of dimensions, wherein the contact information is displayed on the graphical representation at the location, wherein the graphical representation comprises at least a portion of the received content items; receiving, from the user computing device by the host computing device, an indication that a user is finished creating the structural document; generating a print document comprising an encoded data mark; and providing the print document to one or more print-related devices. 3. The method of claim 1 , wherein receiving contact information comprises receiving an email address of the product information provider associated with the structural document. | 0.767016 |
9,497,234 | 8 | 14 | 8. A method comprising: determining a first user interaction represented by an existing social graph connection in a social graph maintained by a social network system; identifying a social network page in the social network system associated with the first user interaction by traversing the social graph of the social network system; identifying a user account based on the first user interaction; generating an implicit social graph connection between the identified user account and the identified social network page based on the determined first user interaction, wherein the existing social graph connection associated with the first user interaction does not involve both the user account and the social network page; determining an edge weight of the implicit social graph connection; storing the implicit social graph connection in the social graph of the social network system; and selecting a content entry for displaying to a user device based on the implicit social graph connection. | 8. A method comprising: determining a first user interaction represented by an existing social graph connection in a social graph maintained by a social network system; identifying a social network page in the social network system associated with the first user interaction by traversing the social graph of the social network system; identifying a user account based on the first user interaction; generating an implicit social graph connection between the identified user account and the identified social network page based on the determined first user interaction, wherein the existing social graph connection associated with the first user interaction does not involve both the user account and the social network page; determining an edge weight of the implicit social graph connection; storing the implicit social graph connection in the social graph of the social network system; and selecting a content entry for displaying to a user device based on the implicit social graph connection. 14. The method of claim 8 , wherein determining the edge weight is based on a history of user interactions related the user account and the social network page. | 0.746032 |
8,849,812 | 1 | 4 | 1. A system, comprising: a processor configured to: determine a topic based on a user demand, wherein the determining of the topic comprises: determine frequencies of phrases based on an analysis of a search query log from a web site, a search query log from a search engine, or a combination thereof; compare the frequencies of the phrases to a predetermined threshold; determine at least one theme based on phrases having frequencies exceeding the predetermined threshold; sort the at least one theme based on common words in the phrases; cluster the at least one theme based on a clustering technique, the clustering technique being a min hash clustering or a n-squared clustering based on bi-grams; and determine the topic based on the clustered at least one theme; automatically generate content for the topic; and select the content that is contextually relevant for display within a corpus of content, wherein the content is optimized for the topic; wherein the corpus of content includes a web site, a user's social networking web page, content customized for mobile devices, content customized based on location awareness, or an electronic mail message; and wherein: in the event that the corpus of content includes the web site, content of the web site is different from other web pages of the website; in the event that the corpus of content includes a user's social networking web page, content of the user's social networking web page is different from another user's social networking web page; in the event that the corpus of content includes content customized for mobile devices, content of a mobile device is different from another mobile device; in the event that the corpus of content includes content customized based on location awareness, content of a location is different from another location; and in the event that the corpus of content includes the electronic mail message, the electronic mail message is different from another electronic mail message; and a memory coupled to the processor and configured to provide the processor with instructions. | 1. A system, comprising: a processor configured to: determine a topic based on a user demand, wherein the determining of the topic comprises: determine frequencies of phrases based on an analysis of a search query log from a web site, a search query log from a search engine, or a combination thereof; compare the frequencies of the phrases to a predetermined threshold; determine at least one theme based on phrases having frequencies exceeding the predetermined threshold; sort the at least one theme based on common words in the phrases; cluster the at least one theme based on a clustering technique, the clustering technique being a min hash clustering or a n-squared clustering based on bi-grams; and determine the topic based on the clustered at least one theme; automatically generate content for the topic; and select the content that is contextually relevant for display within a corpus of content, wherein the content is optimized for the topic; wherein the corpus of content includes a web site, a user's social networking web page, content customized for mobile devices, content customized based on location awareness, or an electronic mail message; and wherein: in the event that the corpus of content includes the web site, content of the web site is different from other web pages of the website; in the event that the corpus of content includes a user's social networking web page, content of the user's social networking web page is different from another user's social networking web page; in the event that the corpus of content includes content customized for mobile devices, content of a mobile device is different from another mobile device; in the event that the corpus of content includes content customized based on location awareness, content of a location is different from another location; and in the event that the corpus of content includes the electronic mail message, the electronic mail message is different from another electronic mail message; and a memory coupled to the processor and configured to provide the processor with instructions. 4. The system recited in claim 1 , wherein select the content that is contextually relevant for display with the corpus of content further comprises determining the content based on a source of the user. | 0.765589 |
9,129,595 | 5 | 6 | 5. The apparatus of claim 4 , wherein the palatometer has in excess of one hundred contact sensors. | 5. The apparatus of claim 4 , wherein the palatometer has in excess of one hundred contact sensors. 6. The apparatus of claim 5 , wherein the palatometer has one hundred and eighteen contact sensors. | 0.5 |
9,373,327 | 13 | 15 | 13. A non-transitory computer readable storage medium having stored thereon instructions that when executed by a processor perform one or more operations, the operations comprising: receiving, using one or more computing devices, a first speech signal corresponding to a first utterance; receiving a second speech signal corresponding to a second utterance, wherein the second utterance is a refinement to the first utterance; determining a first quantity of search results based upon, at least in part, first speech signal information from the first speech signal; determining a second quantity of search results based upon, at least in part, second speech signal information from the second speech signal; comparing at least one of the first quantity of search results and the second quantity of search results with a third quantity of search results; determining an information gain from the comparison; and refining a search based upon at least in part, the information gain. | 13. A non-transitory computer readable storage medium having stored thereon instructions that when executed by a processor perform one or more operations, the operations comprising: receiving, using one or more computing devices, a first speech signal corresponding to a first utterance; receiving a second speech signal corresponding to a second utterance, wherein the second utterance is a refinement to the first utterance; determining a first quantity of search results based upon, at least in part, first speech signal information from the first speech signal; determining a second quantity of search results based upon, at least in part, second speech signal information from the second speech signal; comparing at least one of the first quantity of search results and the second quantity of search results with a third quantity of search results; determining an information gain from the comparison; and refining a search based upon at least in part, the information gain. 15. The computer readable storage medium of claim 13 , further comprising: generating at least one set of search results based upon the refined search. | 0.522152 |
7,639,899 | 5 | 6 | 5. The digital pictorial book system as claimed in claim 4 , wherein said distinguishing feature shows a part of the object, and said informing means informs the user that the part of the main object should be captured. | 5. The digital pictorial book system as claimed in claim 4 , wherein said distinguishing feature shows a part of the object, and said informing means informs the user that the part of the main object should be captured. 6. The digital pictorial book system as claimed in claim 5 , wherein said image database comprises an image of each part of the object as a feature, and said informing means displays the image of the part, which is the distinguishing feature, on said display. | 0.75239 |
7,788,578 | 1 | 3 | 1. A computer-implemented method for modifying appearance and behavior of a markup document having components, comprising: instantiating a web parts page; wherein the web parts page includes web part zones that include one or more web parts; instantiating a tool pane including tool parts for modifying the web parts page by adding code directly within code of the instantiated web parts page such that the instantiation of the tool pane changes a hierarchy of the web parts page; selecting tool parts that provide different functions to include within the tool pane for modifying the web parts page; wherein at least one of the different functions effect the layout of the web parts page; wherein at least another one of the different functions effect the appearance of the web parts page; wherein at least another one of the different functions changes a setting of a web part within the web parts page; and wherein at least another one of the different functions changes a number of web parts included in the web parts page; displaying the tool pane with the selected tool parts as part of the web parts page such that the tool pane is displayed within the web parts page that is modified using one of the provided tool parts; receiving an input to perform one of the functions that is associated with one of the tool parts that is included within the tool pane; wherein performing the function modifies the web parts page; and dynamically modifying the display of the web parts page and the tool pane in response to the modification of the web parts page such that different tool parts are included within the tool pane within the web parts page; wherein the same web parts page remains displayed. | 1. A computer-implemented method for modifying appearance and behavior of a markup document having components, comprising: instantiating a web parts page; wherein the web parts page includes web part zones that include one or more web parts; instantiating a tool pane including tool parts for modifying the web parts page by adding code directly within code of the instantiated web parts page such that the instantiation of the tool pane changes a hierarchy of the web parts page; selecting tool parts that provide different functions to include within the tool pane for modifying the web parts page; wherein at least one of the different functions effect the layout of the web parts page; wherein at least another one of the different functions effect the appearance of the web parts page; wherein at least another one of the different functions changes a setting of a web part within the web parts page; and wherein at least another one of the different functions changes a number of web parts included in the web parts page; displaying the tool pane with the selected tool parts as part of the web parts page such that the tool pane is displayed within the web parts page that is modified using one of the provided tool parts; receiving an input to perform one of the functions that is associated with one of the tool parts that is included within the tool pane; wherein performing the function modifies the web parts page; and dynamically modifying the display of the web parts page and the tool pane in response to the modification of the web parts page such that different tool parts are included within the tool pane within the web parts page; wherein the same web parts page remains displayed. 3. The computer-implemented method of claim 1 , wherein instantiating the tool pane further comprises receiving a field value directing the tool pane to be initiated. | 0.558511 |
9,734,519 | 1 | 4 | 1. A system comprising one or more processors and a non-transitory storage medium comprising program logic for execution by the one or more processors, the program logic comprising: a native advertisement injection engine that: generates a single script block for placement at a single location on an HTML document in response to a creation of one or more ad units for the HTML document, the single script block including one or more section codes that correspond to one or more sections of the HTML document; generates a syndication script for obtaining logic and metadata for injecting native advertisements in the HTML document based on the one or more section codes; provides the syndication script in response to a request generated by the single script block; and provides one or more native advertisements in response to an ad call generated by the syndication script, the one or more native advertisements for injection in the one or more sections of the HTML document based on the logic and metadata, wherein the syndication script obtains, from an advertisement server, logic for filtering child nodes within the Document Object Model to include only matching structures for subsequent injections of native advertisements, and obtains metadata for auto formatting data based on content of the HTML document. | 1. A system comprising one or more processors and a non-transitory storage medium comprising program logic for execution by the one or more processors, the program logic comprising: a native advertisement injection engine that: generates a single script block for placement at a single location on an HTML document in response to a creation of one or more ad units for the HTML document, the single script block including one or more section codes that correspond to one or more sections of the HTML document; generates a syndication script for obtaining logic and metadata for injecting native advertisements in the HTML document based on the one or more section codes; provides the syndication script in response to a request generated by the single script block; and provides one or more native advertisements in response to an ad call generated by the syndication script, the one or more native advertisements for injection in the one or more sections of the HTML document based on the logic and metadata, wherein the syndication script obtains, from an advertisement server, logic for filtering child nodes within the Document Object Model to include only matching structures for subsequent injections of native advertisements, and obtains metadata for auto formatting data based on content of the HTML document. 4. The system of claim 1 , wherein the metadata includes a skip parameter for determining locations for subsequent injections of native advertisements. | 0.602632 |
8,635,094 | 1 | 4 | 1. A computer-implemented method for role-based personalization of a collaborative space comprising: obtaining, using a computer hardware system, role-based information for an interacting user that has been defined by an underlying business process model in a workflow; and generating, in the computer hardware system, the collaborative space utilizing the role-based information, including: parsing the workflow to extract a role model; generating a collaborative space domain model from the role model; selecting a plurality of user interface components based upon the role model, wherein mapping rules are defined to transform role information aggregated in the role model to user interface components for incorporation in the collaborative space domain model, and wherein the mapping rules include user-specified mapping rules between a group of workflow tasks and existing user interface components where existing user interface components exist when the collaborative space is created, and suggested mapping rules obtained by segmenting the workflow based upon role-assignment or a control-flow structure where user interface components do not exist when the collaborative space is created; organizing the selected user interface components in the collaborative space; and rendering the collaborative space. | 1. A computer-implemented method for role-based personalization of a collaborative space comprising: obtaining, using a computer hardware system, role-based information for an interacting user that has been defined by an underlying business process model in a workflow; and generating, in the computer hardware system, the collaborative space utilizing the role-based information, including: parsing the workflow to extract a role model; generating a collaborative space domain model from the role model; selecting a plurality of user interface components based upon the role model, wherein mapping rules are defined to transform role information aggregated in the role model to user interface components for incorporation in the collaborative space domain model, and wherein the mapping rules include user-specified mapping rules between a group of workflow tasks and existing user interface components where existing user interface components exist when the collaborative space is created, and suggested mapping rules obtained by segmenting the workflow based upon role-assignment or a control-flow structure where user interface components do not exist when the collaborative space is created; organizing the selected user interface components in the collaborative space; and rendering the collaborative space. 4. The method of claim 1 , wherein segmenting the workflow based upon a control-flow structure is carried out by comparing structural features of the workflow to group tasks in the workflow. | 0.711246 |
8,418,057 | 14 | 15 | 14. A method for displaying text comprising the steps of: providing a text comprised of a plurality of paragraphs each having a plurality of sentences with each sentence having a plurality of words; arranging said text into a plurality of vertically centered and vertically aligned word clusters having a vertical cluster spacing between consecutive word clusters, each word cluster having a plurality of horizontally extending lines having a vertical line spacing between consecutive lines smaller than said vertical cluster spacing with each line having one or more words, and wherein each word cluster is formed of a thought group defined by a plurality of words linked by a commonality and constrained by an estimate of reader apprehension span, and wherein at least one word in the last line of at least one word cluster has a serif larger than a remainder of said words of said at least one word cluster; and displaying said plurality of word clusters on an electronic display medium. | 14. A method for displaying text comprising the steps of: providing a text comprised of a plurality of paragraphs each having a plurality of sentences with each sentence having a plurality of words; arranging said text into a plurality of vertically centered and vertically aligned word clusters having a vertical cluster spacing between consecutive word clusters, each word cluster having a plurality of horizontally extending lines having a vertical line spacing between consecutive lines smaller than said vertical cluster spacing with each line having one or more words, and wherein each word cluster is formed of a thought group defined by a plurality of words linked by a commonality and constrained by an estimate of reader apprehension span, and wherein at least one word in the last line of at least one word cluster has a serif larger than a remainder of said words of said at least one word cluster; and displaying said plurality of word clusters on an electronic display medium. 15. The method of claim 14 , wherein said commonality linking said plurality of words comprising one of said word clusters comprises a name, a noun or verb which relies on a modifier for understanding, or a phrase, and wherein the end of said thought group of the one of said word clusters is defined by a comma, a colon, or a semicolon. | 0.517192 |
10,134,390 | 1 | 6 | 1. An electronic device comprising: a memory configured to store a user pronunciation lexicon; a voice inputter; and a processor configured to: control the voice inputter to receive a user voice, obtain a word corresponding to the user voice using the user pronunciation lexicon, obtain a first pronunciation for a phoneme included in the user voice, compare the first pronunciation and a second pronunciation pre stored for the word, identify a pronunciation pattern by using a result of the comparing, and update the user pronunciation lexicon according to a pronunciation pattern rule obtained based on the pronunciation pattern, wherein the pronunciation pattern rule is a rule that a user repeats according to at least one of a pronunciation habit of the user and a pronunciation characteristic of the user. | 1. An electronic device comprising: a memory configured to store a user pronunciation lexicon; a voice inputter; and a processor configured to: control the voice inputter to receive a user voice, obtain a word corresponding to the user voice using the user pronunciation lexicon, obtain a first pronunciation for a phoneme included in the user voice, compare the first pronunciation and a second pronunciation pre stored for the word, identify a pronunciation pattern by using a result of the comparing, and update the user pronunciation lexicon according to a pronunciation pattern rule obtained based on the pronunciation pattern, wherein the pronunciation pattern rule is a rule that a user repeats according to at least one of a pronunciation habit of the user and a pronunciation characteristic of the user. 6. The electronic device as claimed in claim 1 , wherein the processor is further configured to generate a variation rule for a predetermined word based on the obtained pronunciation pattern rule and update the user pronunciation lexicon based on the generated variation rule. | 0.703863 |
8,301,448 | 1 | 2 | 1. A method for use with an automatic speech recognition system, the method comprising acts of: analyzing content of a body of speech submitted to a structured document to identify a first section of the structured document to which the body of speech is submitted; in response to identifying the first section, loading a grammar and/or language model for use in recognizing the speech in the body submitted to the first section; and performing speech recognition on the speech in the body using the grammar and/or language model. | 1. A method for use with an automatic speech recognition system, the method comprising acts of: analyzing content of a body of speech submitted to a structured document to identify a first section of the structured document to which the body of speech is submitted; in response to identifying the first section, loading a grammar and/or language model for use in recognizing the speech in the body submitted to the first section; and performing speech recognition on the speech in the body using the grammar and/or language model. 2. The method according to claim 1 , wherein the body of speech is submitted by a first user, the loading comprises loading a language model, and the language model is trained using content selected from a group consisting of content from the first user previously submitted to other structured documents and content from one or more other users previously submitted to the first section. | 0.5 |
8,769,491 | 1 | 10 | 1. A non-transitory computer-readable medium storing instructions, the instructions comprising: one or more instructions that, when executed by one or more processors, cause the one or more processors to: obtain scripting language code that references a collection of code for executing a first task and a second task, a first portion of the collection of code being associated with a first annotation, of a plurality of annotations, associated with the first task, and a second portion of the collection of code being associated with a second annotation, of the plurality of annotations, associated with the second task, select, for each of the first task and the second task, one of a plurality of threads based on the plurality of annotations, the first annotation specifying a first type of thread to which the first portion of the collection of code should be dispatched, and the second annotation specifying a second type of thread to which the second portion of the collection of code should be dispatched; dispatch, based on the first annotation specifying the first type of thread, the first task to a first thread, of the plurality of threads, for executing the first task in a scripting language environment; dispatch, based on the second annotation specifying the second type of thread, the second task to a second thread, of the plurality of threads, included in a non-scripting language environment, the first task or the second task not being dispatched to one or more threads, of the plurality of threads, based on a third annotation, of the plurality of annotations, the third annotation specifying an identity of the one or more threads and indicating that the first task or the second task should not be dispatched to the identified one or more threads; and cause an execution of the scripting language code, during the execution of the scripting language code, the first task being executed via the first thread in the scripting language environment and the second task being executed via the second thread in the non-scripting language environment. | 1. A non-transitory computer-readable medium storing instructions, the instructions comprising: one or more instructions that, when executed by one or more processors, cause the one or more processors to: obtain scripting language code that references a collection of code for executing a first task and a second task, a first portion of the collection of code being associated with a first annotation, of a plurality of annotations, associated with the first task, and a second portion of the collection of code being associated with a second annotation, of the plurality of annotations, associated with the second task, select, for each of the first task and the second task, one of a plurality of threads based on the plurality of annotations, the first annotation specifying a first type of thread to which the first portion of the collection of code should be dispatched, and the second annotation specifying a second type of thread to which the second portion of the collection of code should be dispatched; dispatch, based on the first annotation specifying the first type of thread, the first task to a first thread, of the plurality of threads, for executing the first task in a scripting language environment; dispatch, based on the second annotation specifying the second type of thread, the second task to a second thread, of the plurality of threads, included in a non-scripting language environment, the first task or the second task not being dispatched to one or more threads, of the plurality of threads, based on a third annotation, of the plurality of annotations, the third annotation specifying an identity of the one or more threads and indicating that the first task or the second task should not be dispatched to the identified one or more threads; and cause an execution of the scripting language code, during the execution of the scripting language code, the first task being executed via the first thread in the scripting language environment and the second task being executed via the second thread in the non-scripting language environment. 10. The non-transitory computer-readable medium of claim 1 , where the collection of code includes one or more of: a package, a class, a method, or a field. | 0.910653 |
5,384,701 | 1 | 7 | 1. A system for translating phrases from a first language into a second language, comprising: input means for accepting an input phrase in the first language; a store holding a collection of phrases in the second language; characterization means connected to said input means for determining which phrase of the collection corresponds to the input phrase, and to control the output of that phrase; and output means responsive to the characterization means for outputting the determined phrase in the second language; wherein the characterization means comprises means for recognizing in the input phrase the presence of at least one keyword or keyword parts of a predetermined set of keywords or keyword parts, the number of members in the set of keywords being smaller than the number of phrases in the collection, and to select, in dependence on those recognized keywords or keyword parts, a stored phrase from the collection. | 1. A system for translating phrases from a first language into a second language, comprising: input means for accepting an input phrase in the first language; a store holding a collection of phrases in the second language; characterization means connected to said input means for determining which phrase of the collection corresponds to the input phrase, and to control the output of that phrase; and output means responsive to the characterization means for outputting the determined phrase in the second language; wherein the characterization means comprises means for recognizing in the input phrase the presence of at least one keyword or keyword parts of a predetermined set of keywords or keyword parts, the number of members in the set of keywords being smaller than the number of phrases in the collection, and to select, in dependence on those recognized keywords or keyword parts, a stored phrase from the collection. 7. A system as claimed in claim 1 for providing translations from a first language into any one of a plurality of second languages, a collection of phrases in each of said plurality of second languages being provided in a respective store. | 0.757606 |
9,158,995 | 1 | 10 | 1. A method for object localization in an image comprising: for an input image, generating a task-dependent representation of the input image based on relevance scores for an object to be localized, comprising: for each of a plurality of patches of the image, generating a patch-based representation based on low level features extracted from the patch; with a classifier, outputting a relevance score for each patch, based on the respective patch-based representation; and generating the task-dependent representation based on the patch relevance scores; thereafter, identifying at least one similar image from a set of images, based on the task-dependent representation of the input image and task-dependent representations of images in the set of images; and identifying a location of the object in the input image based on an object location annotation for at least one of the at least one similar images identified in the set of images, wherein at least one of the generating of the task-dependent representation, identifying of the at least one similar image, and the identifying a location of the object in the input image is performed with a computer processor. | 1. A method for object localization in an image comprising: for an input image, generating a task-dependent representation of the input image based on relevance scores for an object to be localized, comprising: for each of a plurality of patches of the image, generating a patch-based representation based on low level features extracted from the patch; with a classifier, outputting a relevance score for each patch, based on the respective patch-based representation; and generating the task-dependent representation based on the patch relevance scores; thereafter, identifying at least one similar image from a set of images, based on the task-dependent representation of the input image and task-dependent representations of images in the set of images; and identifying a location of the object in the input image based on an object location annotation for at least one of the at least one similar images identified in the set of images, wherein at least one of the generating of the task-dependent representation, identifying of the at least one similar image, and the identifying a location of the object in the input image is performed with a computer processor. 10. The method of claim 1 , wherein the identifying of the at least one similar image from the set of images comprises computing a linear kernel between the task-dependent representation of the input image and each of the task-dependent representations of a plurality of the images in the set of images. | 0.830157 |
8,494,837 | 1 | 6 | 1. A method for creating or updating parallel corpus in a machine translation system, comprising the steps of: without parallel corpus, translating a test set E from a first collection to a second collection so as to create a set F in the second collection, and translating the set F from the second collection to the first collection so as to create a set E′ in the first collection, wherein differences between E and E′ are determined; computing confidence scores for a translation of each item in the test set E based on a similarity of E and E′; and based on the confidence scores, adding translations to the parallel corpus, wherein the parallel corpus is stored in memory on the machine translation system. | 1. A method for creating or updating parallel corpus in a machine translation system, comprising the steps of: without parallel corpus, translating a test set E from a first collection to a second collection so as to create a set F in the second collection, and translating the set F from the second collection to the first collection so as to create a set E′ in the first collection, wherein differences between E and E′ are determined; computing confidence scores for a translation of each item in the test set E based on a similarity of E and E′; and based on the confidence scores, adding translations to the parallel corpus, wherein the parallel corpus is stored in memory on the machine translation system. 6. The method of claim 1 , wherein the confidence scores are computed using at least one of Bilingual Evaluation Understudy or Translation Edit Rate scoring metrics. | 0.855517 |
7,519,586 | 13 | 15 | 13. A method of searching comprising: receiving a user query from a user; after receiving said user query, performing a search based on said user query to produce results that are ranked, wherein said results comprise references to entities; contacting said entities to determine whether entities in said results desire to change their rank in said results; charging entities that increase their rank, wherein at least a portion of the amount charged to entities that increase their rank is paid to entities that decrease their rank; crediting entities that decrease their rank; and after charging said entities that increase their rank, crediting said entities that decrease their rank, and changing rankings of said results, reporting said results with changed rankings to said user. | 13. A method of searching comprising: receiving a user query from a user; after receiving said user query, performing a search based on said user query to produce results that are ranked, wherein said results comprise references to entities; contacting said entities to determine whether entities in said results desire to change their rank in said results; charging entities that increase their rank, wherein at least a portion of the amount charged to entities that increase their rank is paid to entities that decrease their rank; crediting entities that decrease their rank; and after charging said entities that increase their rank, crediting said entities that decrease their rank, and changing rankings of said results, reporting said results with changed rankings to said user. 15. The method according to claim 13 , wherein said results are limited in number, such that only entities that are produced by said search are provided an opportunity to change their rank. | 0.5 |
6,064,982 | 5 | 21 | 5. A smart configurator for needs assessment and product configuration, comprising: a configuration tool in the form of a series of cascading style sheets, said configuration tool comprising: a module that assists a user in determining hardware needs; a customer needs identification page for assessing a customer's product requirements, in which a customer may select among various statements; a system needs determination page in which hardware requirements are identified; and a recommended system configuration page that provides a recommended system configuration that includes any of product identification, pricing information, various options that have been selected, and sales terms and conditions that may be desired; a browser for progressing through said style sheets during an interactive, off-line customer needs assessment and product configuration session; and means for automatically recommending a system configuration that most nearly meets a customer's needs, based upon the results of said interactive customer product selection session. | 5. A smart configurator for needs assessment and product configuration, comprising: a configuration tool in the form of a series of cascading style sheets, said configuration tool comprising: a module that assists a user in determining hardware needs; a customer needs identification page for assessing a customer's product requirements, in which a customer may select among various statements; a system needs determination page in which hardware requirements are identified; and a recommended system configuration page that provides a recommended system configuration that includes any of product identification, pricing information, various options that have been selected, and sales terms and conditions that may be desired; a browser for progressing through said style sheets during an interactive, off-line customer needs assessment and product configuration session; and means for automatically recommending a system configuration that most nearly meets a customer's needs, based upon the results of said interactive customer product selection session. 21. The smart configurator of claim 5, further comprising: an overview page for providing a navigation tool in which any step in a session may be accessed directly by clicking on that portion of a graphic representation. | 0.676471 |
8,489,538 | 19 | 22 | 19. A method for analyzing a plurality of documents, comprising: receiving the plurality of documents via a computing device; filtering the plurality of documents to produce a subset of the plurality of documents; executing instructions stored in memory, wherein execution of the instructions by a processor generates an initial control set based on random sampling of the subset of the plurality of documents on a rolling load basis; receiving user input from the computing device, the user input based on an identified subject or category; and executing instructions stored in memory, wherein execution of the instructions by a processor: reviews the initial control set to determine at least one seed set parameter associated with the identified subject or category, automatically codes a first portion of the plurality of documents, based on the initial control set and the at least one seed set parameter associated with the identified subject or category, automatically codes a second portion of the plurality of documents resulting from an application of user analysis and an adaptive identification cycle, and adds the coded second portion of the plurality of documents to the initial control set. | 19. A method for analyzing a plurality of documents, comprising: receiving the plurality of documents via a computing device; filtering the plurality of documents to produce a subset of the plurality of documents; executing instructions stored in memory, wherein execution of the instructions by a processor generates an initial control set based on random sampling of the subset of the plurality of documents on a rolling load basis; receiving user input from the computing device, the user input based on an identified subject or category; and executing instructions stored in memory, wherein execution of the instructions by a processor: reviews the initial control set to determine at least one seed set parameter associated with the identified subject or category, automatically codes a first portion of the plurality of documents, based on the initial control set and the at least one seed set parameter associated with the identified subject or category, automatically codes a second portion of the plurality of documents resulting from an application of user analysis and an adaptive identification cycle, and adds the coded second portion of the plurality of documents to the initial control set. 22. The method of claim 19 , further comprising receiving user input from the computing device, the user input comprising a designation corresponding to key documents of the initial control set. | 0.791398 |
8,666,739 | 8 | 13 | 8. A method for estimating a language model weight, the method comprising: receiving a speech feature vector converted from a speech signal, performing a first search by applying a first language model to the received speech feature vector, and outputting a word lattice and a first acoustic score of the word lattice as a continuous speech recognition result; outputting a second acoustic score as a phoneme recognition result by applying an acoustic model to the speech feature vector; comparing the first acoustic score of the continuous speech recognition result with the second acoustic score of the phoneme recognition result; outputting a first language model weight when the first acoustic score of the continuous speech recognition result is better than the second acoustic score of the phoneme recognition result; and performing a second search by applying a second language model weight, which is the same as the output first language model, to the word lattice. | 8. A method for estimating a language model weight, the method comprising: receiving a speech feature vector converted from a speech signal, performing a first search by applying a first language model to the received speech feature vector, and outputting a word lattice and a first acoustic score of the word lattice as a continuous speech recognition result; outputting a second acoustic score as a phoneme recognition result by applying an acoustic model to the speech feature vector; comparing the first acoustic score of the continuous speech recognition result with the second acoustic score of the phoneme recognition result; outputting a first language model weight when the first acoustic score of the continuous speech recognition result is better than the second acoustic score of the phoneme recognition result; and performing a second search by applying a second language model weight, which is the same as the output first language model, to the word lattice. 13. The method of claim 8 , wherein the word lattice is obtained by defining a plurality of word combinations searched by the first search as information about connections between words. | 0.75 |
8,000,972 | 1 | 4 | 1. A television receiver remote controller, comprising in combination: a first storage device storing electronic program guide (EPG) data as an EPG database that relates content to television channels containing said content, the first storage device forming a part of a television receiver device; the remote controller being contained in a remote controller housing, the housing containing: a second storage device; a data interface that receives the EPG database from the first storage device forming a part of the television receiver device and stores the EPG database on the second storage device; a speech interface that receives speech input from a user and produces speech signals therefrom; a natural language speech processor engine that receives the speech signals and translates the speech signals to a query of the EPG database stored on the second storage device; and a processor that receives results of the query from the natural language speech processor, and either conveys the results of the query to a user utilizing a user interface or sends navigation commands to the receiver. | 1. A television receiver remote controller, comprising in combination: a first storage device storing electronic program guide (EPG) data as an EPG database that relates content to television channels containing said content, the first storage device forming a part of a television receiver device; the remote controller being contained in a remote controller housing, the housing containing: a second storage device; a data interface that receives the EPG database from the first storage device forming a part of the television receiver device and stores the EPG database on the second storage device; a speech interface that receives speech input from a user and produces speech signals therefrom; a natural language speech processor engine that receives the speech signals and translates the speech signals to a query of the EPG database stored on the second storage device; and a processor that receives results of the query from the natural language speech processor, and either conveys the results of the query to a user utilizing a user interface or sends navigation commands to the receiver. 4. The receiver remote controller according to claim 1 , further comprising a remote command transmitter that transmits a command from the remote controller, such command comprising a program selection selected by manual or voice input via the user interface as a result of the query. | 0.5 |
5,586,215 | 1 | 10 | 1. A speech recognition system for recognizing utterances belonging to a pre-established set of allowable candidate utterances using acoustic speech signals and selected concomitant dynamic visual facial feature motion between selected facial features associated with acoustic speech generation, comprising: a. an acoustic feature extraction apparatus for converting signals representative of acoustic speech into a corresponding acoustic feature vector set of signals; b. a dynamic visual feature extraction apparatus for converting signals representative of the selected concomitant dynamic visual facial feature motion associated with acoustic speech generation into a corresponding visual feature vector set of signals; and c. a time delay neural network classifying apparatus for generating a conditional probability distribution of the allowable candidate speech utterances by accepting and operating on a set of current and time delayed dynamic acoustic feature and visual feature vector sets respectively supplied by the acoustic and visual feature extraction apparatus. | 1. A speech recognition system for recognizing utterances belonging to a pre-established set of allowable candidate utterances using acoustic speech signals and selected concomitant dynamic visual facial feature motion between selected facial features associated with acoustic speech generation, comprising: a. an acoustic feature extraction apparatus for converting signals representative of acoustic speech into a corresponding acoustic feature vector set of signals; b. a dynamic visual feature extraction apparatus for converting signals representative of the selected concomitant dynamic visual facial feature motion associated with acoustic speech generation into a corresponding visual feature vector set of signals; and c. a time delay neural network classifying apparatus for generating a conditional probability distribution of the allowable candidate speech utterances by accepting and operating on a set of current and time delayed dynamic acoustic feature and visual feature vector sets respectively supplied by the acoustic and visual feature extraction apparatus. 10. The speech recognition system of claim 1 wherein the time delay neural network classifying apparatus comprises: a. a tapped delay line input layer for accepting a sequence of acoustic and visual time-varying feature vectors and for providing a multiplicity of sequential acoustic and visual feature vectors in parallel at the delay line output taps: b. a hidden layer of neural cells coupled to the output taps of the tapped delay line for enhancement of time dependent features; c. a classification layer of neural cells coupled to the output of the hidden layer neural cells for generating a set of output time varying signals each representative of the probability of its corresponding utterance being present; and d. an averaging layer apparatus, connected to the output of the classification layer, for generating a set of time averaged varying outputs, one for each allowable utterance type, representative of a combined conditional probability that the associated utterance was spoken. | 0.5 |
7,512,537 | 1 | 2 | 1. A natural language illustration system comprising: a language processor component that analyzes natural language input that comprises at least one action from a user to determine a logical form of the input and translates the logical form into an output, the output comprising at least one of an actor, action, object, and background, wherein the natural language input comprises a plurality of sentences; and an animation engine that selects at least one image, applies at least one template to the at least one image and generates for each sentence an animation dynamically based on the at least one image and the at least one template, wherein each animation is linked to the sentence for which the animation was generated and conveys the meaning of the sentence, the animation engine arranges the animations in a same order as the sentences are provided in the natural language input and playing the animations in order to simulate a movie, wherein upon the user changing the order of the sentences the animation engine changes the order of the animations to match the changed order of the sentences without re-analyzing the natural language input and without generating new animations. | 1. A natural language illustration system comprising: a language processor component that analyzes natural language input that comprises at least one action from a user to determine a logical form of the input and translates the logical form into an output, the output comprising at least one of an actor, action, object, and background, wherein the natural language input comprises a plurality of sentences; and an animation engine that selects at least one image, applies at least one template to the at least one image and generates for each sentence an animation dynamically based on the at least one image and the at least one template, wherein each animation is linked to the sentence for which the animation was generated and conveys the meaning of the sentence, the animation engine arranges the animations in a same order as the sentences are provided in the natural language input and playing the animations in order to simulate a movie, wherein upon the user changing the order of the sentences the animation engine changes the order of the animations to match the changed order of the sentences without re-analyzing the natural language input and without generating new animations. 2. The system of claim 1 , wherein the language processor component is a natural language processor component that comprises a natural language processor module that processes the natural language input to determine the logical form of the input. | 0.547794 |
8,532,680 | 10 | 12 | 10. A method of compressing a Short Message Service (SMS) message, the method comprising: comparing characters input by a user with a stored sentence table of recommended sentences and determining whether there is at least one recommended sentence among the recommended sentences of the sentence table that includes the input characters based on a result of the comparing; if it is determined that there is the at least one recommended sentence that includes the input characters, displaying the at least one recommended sentence; displaying a recommended sentence selected from among the at least one recommended sentence that is displayed; converting the displayed recommended sentence that is selected into a compression code; and transmitting an SMS message that includes the recommended sentence converted into the compression code, wherein a size of the compression code for the displayed recommended sentence is a predetermined fixed size regardless of the number of words in the at least one recommended sentence, wherein the predetermined fixed size is used to determine whether a character or a sentence in the SMS message is a compression code. | 10. A method of compressing a Short Message Service (SMS) message, the method comprising: comparing characters input by a user with a stored sentence table of recommended sentences and determining whether there is at least one recommended sentence among the recommended sentences of the sentence table that includes the input characters based on a result of the comparing; if it is determined that there is the at least one recommended sentence that includes the input characters, displaying the at least one recommended sentence; displaying a recommended sentence selected from among the at least one recommended sentence that is displayed; converting the displayed recommended sentence that is selected into a compression code; and transmitting an SMS message that includes the recommended sentence converted into the compression code, wherein a size of the compression code for the displayed recommended sentence is a predetermined fixed size regardless of the number of words in the at least one recommended sentence, wherein the predetermined fixed size is used to determine whether a character or a sentence in the SMS message is a compression code. 12. The method of claim 10 , wherein each of the predetermined fixed size is two bytes. | 0.928454 |
8,180,640 | 6 | 13 | 6. A computer-implemented method, the method, performed by one or more computers comprising a processor and storage coupled therewith, the method comprising: receiving first acoustic data having been captured from a speaker when speaking a name and storing the first acoustic data in the storage; performing, by the processor, recognition on the first acoustic data to determine that the first acoustic data corresponds to a first name; receiving first interactive input confirming whether the first name is correct, and: when the first interactive input confirms that the first name is correct, using, by the processor, the first name as a grapheme label for the first acoustic data, and the processor capturing and storing in the storage second acoustic data when the speaker speaks the name, recognizing a second name, receiving interactive input inputted to the one or more computers and confirming whether the second name is correct, and: when the second interactive input confirms the second name, using the second name as a grapheme label for the first or second acoustic data, and when the second interactive input does not confirm the second name, receiving an interactively inputted spelling of the name and using the spelling of the name as the grapheme label for the first or second acoustic data; and using, by the processor, the acoustic data and associated grapheme label for subsequent speech recognition, wherein using the acoustic data and associated grapheme label for subsequent speech recognition comprises retraining a model that maps between labeled grapheme sequences and phoneme sequences based on the acoustic data and the associated grapheme label. | 6. A computer-implemented method, the method, performed by one or more computers comprising a processor and storage coupled therewith, the method comprising: receiving first acoustic data having been captured from a speaker when speaking a name and storing the first acoustic data in the storage; performing, by the processor, recognition on the first acoustic data to determine that the first acoustic data corresponds to a first name; receiving first interactive input confirming whether the first name is correct, and: when the first interactive input confirms that the first name is correct, using, by the processor, the first name as a grapheme label for the first acoustic data, and the processor capturing and storing in the storage second acoustic data when the speaker speaks the name, recognizing a second name, receiving interactive input inputted to the one or more computers and confirming whether the second name is correct, and: when the second interactive input confirms the second name, using the second name as a grapheme label for the first or second acoustic data, and when the second interactive input does not confirm the second name, receiving an interactively inputted spelling of the name and using the spelling of the name as the grapheme label for the first or second acoustic data; and using, by the processor, the acoustic data and associated grapheme label for subsequent speech recognition, wherein using the acoustic data and associated grapheme label for subsequent speech recognition comprises retraining a model that maps between labeled grapheme sequences and phoneme sequences based on the acoustic data and the associated grapheme label. 13. A method according to claim 6 , further comprising repeatedly applying a convergence algorithm. | 0.674342 |
7,613,690 | 1 | 9 | 1. A method, performed at least in part by a computer, for identifying a reason that a search topic is popular, the method comprising: receiving an indication of a search topic that is popular, the search topic that is popular being related to a particular entity; in response to receiving the indication of the search topic that is popular, identifying a content feed from an electronic source of published information that includes content published less than a threshold period of time prior to the search topic becoming popular, the content of the content feed including metadata; determining whether the content, of the identified content feed, published less than the threshold period of time prior to the search topic becoming popular is relevant to the search topic; in response to a determination that the content, of the identified content feed, published less than the threshold period of time prior to the search topic becoming popular is relevant to the search topic: analyzing the content, of the identified content feed, published less than the threshold period of time prior to the search topic becoming popular, analyzing the content comprising determining that the content and metadata of the content feed are related to the same particular entity as the search topic; determining a reason that the search topic is popular based on the analysis, determining the reason comprising, when the content identifies the search topic, providing at least some of the content as the reason that the search topic is popular; and presenting to a user the search topic that is popular and the determined reason that the search topic is popular; and summarizing content of more than one content feed when the content of more than one content feed relates to the same particular entity as the search topic. | 1. A method, performed at least in part by a computer, for identifying a reason that a search topic is popular, the method comprising: receiving an indication of a search topic that is popular, the search topic that is popular being related to a particular entity; in response to receiving the indication of the search topic that is popular, identifying a content feed from an electronic source of published information that includes content published less than a threshold period of time prior to the search topic becoming popular, the content of the content feed including metadata; determining whether the content, of the identified content feed, published less than the threshold period of time prior to the search topic becoming popular is relevant to the search topic; in response to a determination that the content, of the identified content feed, published less than the threshold period of time prior to the search topic becoming popular is relevant to the search topic: analyzing the content, of the identified content feed, published less than the threshold period of time prior to the search topic becoming popular, analyzing the content comprising determining that the content and metadata of the content feed are related to the same particular entity as the search topic; determining a reason that the search topic is popular based on the analysis, determining the reason comprising, when the content identifies the search topic, providing at least some of the content as the reason that the search topic is popular; and presenting to a user the search topic that is popular and the determined reason that the search topic is popular; and summarizing content of more than one content feed when the content of more than one content feed relates to the same particular entity as the search topic. 9. The method of claim 1 wherein the content feed from an electronic source of published information includes data in the form of at least one of audio, video, text, audible text after a text-to-speech conversion, images and animation. | 0.914608 |
8,824,785 | 29 | 30 | 29. The system of claim 20 wherein the characteristic of each of the plurality of pixels substantially comprises a color of a plurality of colors. | 29. The system of claim 20 wherein the characteristic of each of the plurality of pixels substantially comprises a color of a plurality of colors. 30. The system of claim 29 further comprising: the processor further operative to filter out pixels of the plurality of pixels from the electronic document image if the color of the pixels of the plurality of pixels comprises a shade of black or a shade of grey. | 0.5 |
8,346,683 | 1 | 8 | 1. A computer system tangibly operating in an information technology hardware and software environment, comprising: a knowledge base model for representation and storage of regulatory knowledge, the regulatory knowledge including the following knowledge base model entities: regulations, procedures, parameters, concepts, decisions, rules, service calls, and their features and relationships to other knowledge base model entities; a data interface model for representation and storage of regulatory data, the regulatory data including the following groups of data interface model entities: simple transactional events, complex events, referential entities, profiles, and their features and their relationships to other data interface model entities and their relationships to knowledge base model entities; a reasoning session data model for representation and storage of the reasoning session data, the reasoning session data including the following reasoning session data model entities: reasoning sessions, session events, and their features and their relationships to other session data model entities and their relationships to data interface model entities and their relationships to knowledge base model entities; an interface configured to receive a request in the form of a structured or semi-structured message from an external source, to pass the received message to a reasoning session controller, and upon receiving a response from the reasoning session controller, pass that response to the external source; a reasoning session controller configured to receive an input request from the interface and to match the input request to a decision record from the knowledge base model decision entity, execute programmable instructions from the decision record, perform post-processing tasks after execution of the programmable instructions is completed, and select a next new session event for execution; a library of procedures configured to be invoked by a service call, take a subset of parameters established by a reasoning session and use those parameters to call external services, and place results returned by the external services in a data interface repository. | 1. A computer system tangibly operating in an information technology hardware and software environment, comprising: a knowledge base model for representation and storage of regulatory knowledge, the regulatory knowledge including the following knowledge base model entities: regulations, procedures, parameters, concepts, decisions, rules, service calls, and their features and relationships to other knowledge base model entities; a data interface model for representation and storage of regulatory data, the regulatory data including the following groups of data interface model entities: simple transactional events, complex events, referential entities, profiles, and their features and their relationships to other data interface model entities and their relationships to knowledge base model entities; a reasoning session data model for representation and storage of the reasoning session data, the reasoning session data including the following reasoning session data model entities: reasoning sessions, session events, and their features and their relationships to other session data model entities and their relationships to data interface model entities and their relationships to knowledge base model entities; an interface configured to receive a request in the form of a structured or semi-structured message from an external source, to pass the received message to a reasoning session controller, and upon receiving a response from the reasoning session controller, pass that response to the external source; a reasoning session controller configured to receive an input request from the interface and to match the input request to a decision record from the knowledge base model decision entity, execute programmable instructions from the decision record, perform post-processing tasks after execution of the programmable instructions is completed, and select a next new session event for execution; a library of procedures configured to be invoked by a service call, take a subset of parameters established by a reasoning session and use those parameters to call external services, and place results returned by the external services in a data interface repository. 8. The system of claim 1 , wherein the rule entity comprises zero or more conditions expressed in a conjunctive normal form, each conjunct of the conjunctive normal form being represented by a feature-operator-value triple wherein a feature name of the feature-operator-value triple corresponds to a feature name from the data interface model or a derived feature name, and the feature value is in a form that is one of numeric, Boolean, date, relation to a concept or relation to a parameter entity; and zero or more actions represented as feature-value pairs wherein the value can be assigned or computed, or as relations to concept entities. | 0.536691 |
4,853,953 | 3 | 4 | 3. A voice controlled automatic dialer as claimed in claim 2, wherein said sound generating means comprises a tone generator for generating a multifrequency tone burst in response to the entry of an input utterance to said sound input means when said input utterance is a dialing number. | 3. A voice controlled automatic dialer as claimed in claim 2, wherein said sound generating means comprises a tone generator for generating a multifrequency tone burst in response to the entry of an input utterance to said sound input means when said input utterance is a dialing number. 4. A voice controlled automatic dialer as claimed in claim 3, wherein said tone generator generates a single frequency tone burst in response to the entry of said input utterance when same is a command word. | 0.5 |
8,374,866 | 17 | 20 | 17. A computer-implemented method comprising: obtaining, by a computer system, aligned voice data that comprises untransformed voice data in the target language that has been aligned with phonemes in a corresponding textual transcript using a source acoustic model for a source language, wherein the source acoustic model includes information that maps acoustic features of the source language to phonemes in a transformed feature space, and the untransformed voice data is in an untransformed feature space; transforming the aligned voice data according to a particular transform operation using the source acoustic model to obtain transformed voice data; adapting the source acoustic model to the target language using the untransformed voice data in the target language to obtain an adapted acoustic model; and training, by the computer system, a target acoustic model for the target language using the transformed voice data and the adapted acoustic model; and providing the target acoustic model in association with the target language. | 17. A computer-implemented method comprising: obtaining, by a computer system, aligned voice data that comprises untransformed voice data in the target language that has been aligned with phonemes in a corresponding textual transcript using a source acoustic model for a source language, wherein the source acoustic model includes information that maps acoustic features of the source language to phonemes in a transformed feature space, and the untransformed voice data is in an untransformed feature space; transforming the aligned voice data according to a particular transform operation using the source acoustic model to obtain transformed voice data; adapting the source acoustic model to the target language using the untransformed voice data in the target language to obtain an adapted acoustic model; and training, by the computer system, a target acoustic model for the target language using the transformed voice data and the adapted acoustic model; and providing the target acoustic model in association with the target language. 20. The computer-implemented method of claim 17 , wherein the particular transform operation comprises a VTLN transform operation that is performed on the aligned voice data using the source acoustic model. | 0.752998 |
7,475,337 | 11 | 12 | 11. A method for producing a structured document, the method comprising: activating an environment including a first display and a second display, the first display displaying an output presentation and the second display displaying a definition file including document type definitions (DTD), wherein the output presentation including a number of displayable objects being displayed and respective decoration attributes about each of the displayable objects, wherein each of the document type definitions includes an identifier; forming a number of group objects, each of the group objects including one or more of the displayable objects; and associating each of the group objects with the identifier in one of the document type definitions; and creating the structured document from the output presentation in accordance with the at least one of the definitions being associated with the one of the displayable objects. | 11. A method for producing a structured document, the method comprising: activating an environment including a first display and a second display, the first display displaying an output presentation and the second display displaying a definition file including document type definitions (DTD), wherein the output presentation including a number of displayable objects being displayed and respective decoration attributes about each of the displayable objects, wherein each of the document type definitions includes an identifier; forming a number of group objects, each of the group objects including one or more of the displayable objects; and associating each of the group objects with the identifier in one of the document type definitions; and creating the structured document from the output presentation in accordance with the at least one of the definitions being associated with the one of the displayable objects. 12. The method of claim 11 further comprising generating a modified output presentation including information of each of the group objects being associated with the identifier in one of the document type definitions. | 0.5 |
7,786,994 | 1 | 2 | 1. A machine-readable storage medium comprising machine-readable instructions that, when executed by the machine, cause the machine to perform a method comprising: receiving an indication of at least one command relating to text that contains at least one first glyph; determining that the at least one first glyph maps to multiple corresponding Unicode representations, the multiple corresponding Unicode representations defining a collision; resolving the collision such that the at least one first glyph maps to only a single Unicode representation of the multiple corresponding Unicode representations; converting the at least one first glyph to the single corresponding Unicode representation in response to the command; performing the command on the single Unicode representation of the at least one first glyph; creating at least one further glyph for the single Unicode representation of the at least one first glyph; and mapping the at least one further glyph element to the at least one first glyph via a glyph substitution table, the glyph substitution table having data structures to enable mapping via: a) single substitution; b) multiple substitution; c) alternative substitution; d) ligature substitution; and e) context substitution. | 1. A machine-readable storage medium comprising machine-readable instructions that, when executed by the machine, cause the machine to perform a method comprising: receiving an indication of at least one command relating to text that contains at least one first glyph; determining that the at least one first glyph maps to multiple corresponding Unicode representations, the multiple corresponding Unicode representations defining a collision; resolving the collision such that the at least one first glyph maps to only a single Unicode representation of the multiple corresponding Unicode representations; converting the at least one first glyph to the single corresponding Unicode representation in response to the command; performing the command on the single Unicode representation of the at least one first glyph; creating at least one further glyph for the single Unicode representation of the at least one first glyph; and mapping the at least one further glyph element to the at least one first glyph via a glyph substitution table, the glyph substitution table having data structures to enable mapping via: a) single substitution; b) multiple substitution; c) alternative substitution; d) ligature substitution; and e) context substitution. 2. The machine-readable storage medium of claim 1 , wherein the instructions for converting the glyph include instructions for reverse mapping the glyph through a character mapping table. | 0.644487 |
7,792,838 | 18 | 19 | 18. The method of claim 17 , wherein an instance is considered to belong to a virtual class of size one and a description for the instance is defined in terms of membership of the instance to the virtual class of size one. | 18. The method of claim 17 , wherein an instance is considered to belong to a virtual class of size one and a description for the instance is defined in terms of membership of the instance to the virtual class of size one. 19. The method of claim 18 , wherein the virtual class is associated with a taxonomy defined based on the domain of the at least one given ontology property. | 0.5 |
9,692,778 | 1 | 5 | 1. A method for prioritizing vulnerabilities of a specific asset deployed by an organization in a specific virtual computing environment, performed by a processor-based contextual vulnerabilities prioritization system, comprising: determining a vulnerability score for the specific asset, based on a CVSS (common vulnerability scoring system) score or other base vulnerability score or temporal vulnerability score, wherein the specific asset is a virtual machine or virtual application that is implemented using physical computing components in the specific virtual computing environment; receiving information about a threat; correlating the information about the threat with information about the specific asset based upon environmental factors of the specific asset to determine a threat score for the specific asset, wherein the environmental factors include characteristics of a customer associated with the specific asset, characteristics of the specific asset relative to the threat, and characteristics of a workload distribution relative to the threat; determining a contextual score for the specific asset based on at least one tag of the specific asset; and deriving a prioritization score for the specific asset, the prioritization score a combination of the vulnerability score, the threat score and the contextual score, the prioritization score representing a prioritizing, specific to the specific asset, of a context-dependent vulnerability of the specific asset to the threat. | 1. A method for prioritizing vulnerabilities of a specific asset deployed by an organization in a specific virtual computing environment, performed by a processor-based contextual vulnerabilities prioritization system, comprising: determining a vulnerability score for the specific asset, based on a CVSS (common vulnerability scoring system) score or other base vulnerability score or temporal vulnerability score, wherein the specific asset is a virtual machine or virtual application that is implemented using physical computing components in the specific virtual computing environment; receiving information about a threat; correlating the information about the threat with information about the specific asset based upon environmental factors of the specific asset to determine a threat score for the specific asset, wherein the environmental factors include characteristics of a customer associated with the specific asset, characteristics of the specific asset relative to the threat, and characteristics of a workload distribution relative to the threat; determining a contextual score for the specific asset based on at least one tag of the specific asset; and deriving a prioritization score for the specific asset, the prioritization score a combination of the vulnerability score, the threat score and the contextual score, the prioritization score representing a prioritizing, specific to the specific asset, of a context-dependent vulnerability of the specific asset to the threat. 5. The method of claim 1 , wherein the contextual score is based on one of: whether the specific asset has sensitive data, whether the specific asset has a critical server, whether the specific asset is Web-connected, whether there is a virus found in the specific asset, whether there is a detected intrusion to the specific asset, whether there is suspect data transferred to the specific asset, or whether there is suspect data transferred from the specific asset. | 0.686156 |
8,478,792 | 7 | 24 | 7. The method of claim 1 wherein the first one of the plurality of grouping labels is extracted from a contextual search code block included in the first content item. | 7. The method of claim 1 wherein the first one of the plurality of grouping labels is extracted from a contextual search code block included in the first content item. 24. A computer-readable storage medium carrying one or more sequences of instructions which, when executed by one or more processors, causes the one or more processors to perform the method recited in claim 7 . | 0.5 |
4,380,069 | 2 | 4 | 2. An apparatus for handling data code words, said apparatus comprising detecting means for detecting erroneous code words, estimating means coupled to said detecting means for estimating replacement code words both for said detected erroneous code words and for at least some code words within a selected interval of said erroneous words, and replacing means coupled to said detecting means and said estimating means for replacing said erroneous code word and said code words within said selected interval by said estimated code words respectively, whereby most undetected errors are thus concealed. | 2. An apparatus for handling data code words, said apparatus comprising detecting means for detecting erroneous code words, estimating means coupled to said detecting means for estimating replacement code words both for said detected erroneous code words and for at least some code words within a selected interval of said erroneous words, and replacing means coupled to said detecting means and said estimating means for replacing said erroneous code word and said code words within said selected interval by said estimated code words respectively, whereby most undetected errors are thus concealed. 4. An apparatus as claimed in claim 2, wherein said detecting means comprises a read only memory. | 0.826786 |
7,856,375 | 87 | 102 | 87. A system for preparing personalized communication documents for a plurality of consumer entities comprising: a computing system; one or more electronic databases coupled to the computing system; one or more software routines executing on the computing system which are adapted to: i) process data from a first electronic data file in said one or more electronic databases containing financial product and/or financial service data for the customized communications, which financial product and/or financial service data includes a plurality of separate descriptions, characteristics and/or identifications for at least a first financial product and/or financial service; ii) process data from a second electronic data file in said one or more electronic databases containing customer information for certain of the plurality of consumer entities, said customer information including customer related data in addition to, but not excluding, any one or more of customer name, customer address and customer account information obtained for said certain of the plurality of consumer entities; iii) perform an automated composition process to compose an electronic document file representing a customized communication document for at least one of said certain of the plurality of the consumer entities, said customized communication document comprising information relating to an offering for one or more financial products or services; wherein at least some content included in said customized communication document is customized content generated for said electronic file which includes variable data specifically for a consumer entity automatically derived and/or calculated from said first electronic data file and/or said second electronic data file during said automated composition process for said electronic file and said consumer entity. | 87. A system for preparing personalized communication documents for a plurality of consumer entities comprising: a computing system; one or more electronic databases coupled to the computing system; one or more software routines executing on the computing system which are adapted to: i) process data from a first electronic data file in said one or more electronic databases containing financial product and/or financial service data for the customized communications, which financial product and/or financial service data includes a plurality of separate descriptions, characteristics and/or identifications for at least a first financial product and/or financial service; ii) process data from a second electronic data file in said one or more electronic databases containing customer information for certain of the plurality of consumer entities, said customer information including customer related data in addition to, but not excluding, any one or more of customer name, customer address and customer account information obtained for said certain of the plurality of consumer entities; iii) perform an automated composition process to compose an electronic document file representing a customized communication document for at least one of said certain of the plurality of the consumer entities, said customized communication document comprising information relating to an offering for one or more financial products or services; wherein at least some content included in said customized communication document is customized content generated for said electronic file which includes variable data specifically for a consumer entity automatically derived and/or calculated from said first electronic data file and/or said second electronic data file during said automated composition process for said electronic file and said consumer entity. 102. The system of claim 87 , where each customized communication document is included along with a host document. | 0.884381 |
8,869,106 | 2 | 3 | 2. The computer program product in accordance with claim 1 , wherein when the language service provider port component holds a plurality of language service providers, the management component is further configured to select one of the plurality of language service providers to provide the set of available symbols. | 2. The computer program product in accordance with claim 1 , wherein when the language service provider port component holds a plurality of language service providers, the management component is further configured to select one of the plurality of language service providers to provide the set of available symbols. 3. The computer program product in accordance with claim 2 , wherein the management component is configured to select the language service provider based at least in part on a user preference. | 0.744 |
10,120,929 | 22 | 28 | 22. A computer-readable medium having instructions stored thereon, wherein the instructions, when executed by a computing apparatus, cause the computing apparatus to: receive item information associated with an item of interest, the item information comprising one or more item attributes and corresponding item attribute values; assign a first category for the item of interest from hierarchically organized first categories stored by a data store; determine one or more rules associated with the assigned first category; identify one or more hierarchically organized second category candidates using the one or more rules and relevance values associated with the item information and each of the second category candidates; determine which of the second category candidates has the most second category candidate attributes of the second category candidates; and select the second category candidate that has the most second category candidate attributes as the second category of the item of interest. | 22. A computer-readable medium having instructions stored thereon, wherein the instructions, when executed by a computing apparatus, cause the computing apparatus to: receive item information associated with an item of interest, the item information comprising one or more item attributes and corresponding item attribute values; assign a first category for the item of interest from hierarchically organized first categories stored by a data store; determine one or more rules associated with the assigned first category; identify one or more hierarchically organized second category candidates using the one or more rules and relevance values associated with the item information and each of the second category candidates; determine which of the second category candidates has the most second category candidate attributes of the second category candidates; and select the second category candidate that has the most second category candidate attributes as the second category of the item of interest. 28. The computer-readable medium of claim 22 , further comprising instructions that cause the computing apparatus to update the item information with the assigned first category and the assigned second category and to transmit the updated item information to a data store. | 0.695749 |
8,902,287 | 1 | 6 | 1. A method for displaying a three-dimensional (3D) caption in a 3D display apparatus, the method comprising: receiving a broadcast signal including 3D caption data based on a code space, wherein the code space contains base code sets and extended code sets; acquiring 3D command and caption text from the 3D caption data, wherein the 3D command is delivered in at least one extended code set, wherein the at least one extended code set is accessed by using an ‘EXT1’ code in a base code set, wherein the 3D command provides 3D information of a caption window; and processing the 3D information and the caption text such that the caption text is written to the caption window for 3D captioning. | 1. A method for displaying a three-dimensional (3D) caption in a 3D display apparatus, the method comprising: receiving a broadcast signal including 3D caption data based on a code space, wherein the code space contains base code sets and extended code sets; acquiring 3D command and caption text from the 3D caption data, wherein the 3D command is delivered in at least one extended code set, wherein the at least one extended code set is accessed by using an ‘EXT1’ code in a base code set, wherein the 3D command provides 3D information of a caption window; and processing the 3D information and the caption text such that the caption text is written to the caption window for 3D captioning. 6. The method of claim 1 , wherein the caption text is three-dimensionally displayed within the caption window. | 0.74424 |
8,949,223 | 10 | 16 | 10. The method according to claim 1 , further comprising the steps of: identifying, among a plurality of translatable components, at least two associated translatable components; and generating one or more translated components in the second language that are locked-together corresponding to the at least two associated translatable components in the first language. | 10. The method according to claim 1 , further comprising the steps of: identifying, among a plurality of translatable components, at least two associated translatable components; and generating one or more translated components in the second language that are locked-together corresponding to the at least two associated translatable components in the first language. 16. The method according to claim 10 , wherein the step of generating the content in the second language is conditioned by the following steps of: determining a percentage of the content in the first language for which the corresponding translated previous content in the second language is stored; and conditioning the step of generating the content in the second language based on the percentage so determined. | 0.5 |
9,129,214 | 1 | 9 | 1. A method comprising: receiving title interaction data, wherein the title interaction data specifies, for each user of a plurality of users, an order in which the user interacted with a plurality of titles; generating a plurality of statistical models, each statistical model of the plurality of statistical models specifying a plurality of probabilities, wherein the plurality of probabilities represent, for each first title of the plurality of titles and each second title of the plurality of titles, a likelihood that a user will interact with the first title then next interact with the second title; wherein generating the plurality of statistical models is performed by creating a global statistical model based on the title interaction data and applying noise to the global statistical model to create each of the plurality of statistical models; refining the plurality of statistical models based on the title interaction data to produce a plurality of refined statistical models; determining a plurality of weight values corresponding to the plurality of refined statistical models for a particular user of the plurality of users, wherein each weight value of the plurality of weight values corresponds to a respective refined statistical model of the plurality of refined statistical models; identifying, for the particular user, one or more recommended titles of the plurality of titles based on the plurality of weight values and the plurality of refined statistical models; wherein the method is performed by one or more computing devices. | 1. A method comprising: receiving title interaction data, wherein the title interaction data specifies, for each user of a plurality of users, an order in which the user interacted with a plurality of titles; generating a plurality of statistical models, each statistical model of the plurality of statistical models specifying a plurality of probabilities, wherein the plurality of probabilities represent, for each first title of the plurality of titles and each second title of the plurality of titles, a likelihood that a user will interact with the first title then next interact with the second title; wherein generating the plurality of statistical models is performed by creating a global statistical model based on the title interaction data and applying noise to the global statistical model to create each of the plurality of statistical models; refining the plurality of statistical models based on the title interaction data to produce a plurality of refined statistical models; determining a plurality of weight values corresponding to the plurality of refined statistical models for a particular user of the plurality of users, wherein each weight value of the plurality of weight values corresponds to a respective refined statistical model of the plurality of refined statistical models; identifying, for the particular user, one or more recommended titles of the plurality of titles based on the plurality of weight values and the plurality of refined statistical models; wherein the method is performed by one or more computing devices. 9. The method of claim 1 , further comprising receiving data indicating that the particular user has currently interacted with a particular title of the plurality of titles, wherein the one or more recommended titles are identified based, at least in part, on the particular title. | 0.59157 |
8,386,396 | 1 | 13 | 1. A computer implemented method for bidirectional matching between a plurality of parties belonging to a first group and a plurality of parties belonging to a second group, the method comprising: (a) operating a processor thereby to collect, from each party belonging to the first group, data indicative of first group criterion responses for a plurality of criterions, wherein each criterion has a pre-assigned baseline rating score; (b) operating the processor thereby to determine, in respect of at least a selection of the first group criterion responses, first group preferential rating scores; (c) operating the processor thereby to collect, from each party belonging to the second group, data indicative of second group criterion responses for the plurality of criterions; (d) operating the processor thereby to determine, in respect of at least a selection of the second group criterion responses, second group preferential rating scores; (e) on the basis of a criterion match determination protocol, operating the processor thereby to process the data indicative of criterion responses, thereby to identify criterion matches between parties belonging to the first group and parties belonging to the second group; (f) operating the processor thereby to determine, for each criterion match between a party belonging to the first group and a party belonging to the second group, a criterion match rating based on a function of the baseline rating score and, where determined, the first group preferential rating score, the second group preferential rating score, or the first group preferential rating score and the second group preferential rating score in combination; and (g) operating the processor thereby to calculate, in respect of a match between a given party belonging to the first group and a given party belonging to the second group, a total match rating, wherein the total match rating is calculated based on criterion match ratings determined for criterion matches between the given party belonging to the first group and the given party belonging to the second group, including at least one criterion match rating based on a first group preferential rating score and at least one criterion match rating based on a second group preferential rating score. | 1. A computer implemented method for bidirectional matching between a plurality of parties belonging to a first group and a plurality of parties belonging to a second group, the method comprising: (a) operating a processor thereby to collect, from each party belonging to the first group, data indicative of first group criterion responses for a plurality of criterions, wherein each criterion has a pre-assigned baseline rating score; (b) operating the processor thereby to determine, in respect of at least a selection of the first group criterion responses, first group preferential rating scores; (c) operating the processor thereby to collect, from each party belonging to the second group, data indicative of second group criterion responses for the plurality of criterions; (d) operating the processor thereby to determine, in respect of at least a selection of the second group criterion responses, second group preferential rating scores; (e) on the basis of a criterion match determination protocol, operating the processor thereby to process the data indicative of criterion responses, thereby to identify criterion matches between parties belonging to the first group and parties belonging to the second group; (f) operating the processor thereby to determine, for each criterion match between a party belonging to the first group and a party belonging to the second group, a criterion match rating based on a function of the baseline rating score and, where determined, the first group preferential rating score, the second group preferential rating score, or the first group preferential rating score and the second group preferential rating score in combination; and (g) operating the processor thereby to calculate, in respect of a match between a given party belonging to the first group and a given party belonging to the second group, a total match rating, wherein the total match rating is calculated based on criterion match ratings determined for criterion matches between the given party belonging to the first group and the given party belonging to the second group, including at least one criterion match rating based on a first group preferential rating score and at least one criterion match rating based on a second group preferential rating score. 13. A method according to claim 1 wherein the baseline rating scores are relative between criterions across the decision set and wherein the preferential rating scores are relative with respect to a predefined frame of reference. | 0.556202 |
9,245,037 | 12 | 15 | 12. A non-transitory computer-readable medium storing computer-readable instructions that, when executed by a computing device, cause the computing device to: receive a first request to analyze a first network document, the first request including an identifier of the first network document; in response to the first request: analyze the first network document according to one or more search engine algorithms; generate a display of a first scoring analysis including a first score of the first network document; and generate a display of an option to view a second scoring analysis of a second network document within the first network document, wherein the second network document includes at least one link contributing to the first scoring analysis; in response to receiving a selection of the option to view the second scoring analysis, generate a display of results of the second scoring analysis including a second score of the second network document, wherein the display of the results of the second scoring analysis includes a link flow distribution that indicates a likelihood that a user will access the second network document relative to a third network document within the first network document, and wherein the display of the results of the first scoring analysis includes at least one factor contributing to the first score and wherein the display of the results of the second scoring analysis does not include the at least one factor contributing to the first score; receive a second request to view a third scoring analysis of the at least one link, wherein the third scoring analysis includes an evaluation of one or more traffic-independent attributes of the at least one link, wherein the one or more traffic-independent attributes of the at least one link is different from attributes of a network destination specified by the at least one link; and in response to receiving the second request, generate a display of results of the third scoring analysis of the at least one link including the one or more traffic-independent attributes. | 12. A non-transitory computer-readable medium storing computer-readable instructions that, when executed by a computing device, cause the computing device to: receive a first request to analyze a first network document, the first request including an identifier of the first network document; in response to the first request: analyze the first network document according to one or more search engine algorithms; generate a display of a first scoring analysis including a first score of the first network document; and generate a display of an option to view a second scoring analysis of a second network document within the first network document, wherein the second network document includes at least one link contributing to the first scoring analysis; in response to receiving a selection of the option to view the second scoring analysis, generate a display of results of the second scoring analysis including a second score of the second network document, wherein the display of the results of the second scoring analysis includes a link flow distribution that indicates a likelihood that a user will access the second network document relative to a third network document within the first network document, and wherein the display of the results of the first scoring analysis includes at least one factor contributing to the first score and wherein the display of the results of the second scoring analysis does not include the at least one factor contributing to the first score; receive a second request to view a third scoring analysis of the at least one link, wherein the third scoring analysis includes an evaluation of one or more traffic-independent attributes of the at least one link, wherein the one or more traffic-independent attributes of the at least one link is different from attributes of a network destination specified by the at least one link; and in response to receiving the second request, generate a display of results of the third scoring analysis of the at least one link including the one or more traffic-independent attributes. 15. The non-transitory computer-readable medium of claim 12 , wherein the first scoring analysis includes analyzing link statistics. | 0.876404 |
7,644,209 | 12 | 13 | 12. The handheld electronic device of claim 11 wherein the memory has stored therein a map file comprising an assignment of each linguistic element to a corresponding input member, at least some of the linguistic elements in the map file being in the alphabet. | 12. The handheld electronic device of claim 11 wherein the memory has stored therein a map file comprising an assignment of each linguistic element to a corresponding input member, at least some of the linguistic elements in the map file being in the alphabet. 13. The handheld electronic device of claim 12 wherein, responsive to detecting a predetermined input comprising an actuation of a particular input member, the processor apparatus is adapted to output at least a portion of a set of the linguistic elements from the map file that are assigned to the particular input member. | 0.5 |
8,185,813 | 1 | 4 | 1. A method comprising: performing, by a computer system, an automated keyword analysis of an electronic document to identify a set of keywords associated with the electronic document; displaying, by the computer system, the set of keywords; accepting, by the computer system, user input indicating user-specified concepts of interest, wherein the user input includes at least one keyword identified by the automated keyword analysis and at least one keyword not identified by the automated keyword analysis; analyzing, by the computer system, the electronic document to identify locations of discussion of the user-specified concepts of interest; and displaying, by the computer system, a graph representing the electronic document and illustrating persistence values associated with the user-specified concepts of interest at locations in the electronic document, wherein, for a given location in the electronic document, the graph illustrates a persistence value indicating a frequency of discussion of a user-specified concept of interest at that location relative to other locations in the electronic document. | 1. A method comprising: performing, by a computer system, an automated keyword analysis of an electronic document to identify a set of keywords associated with the electronic document; displaying, by the computer system, the set of keywords; accepting, by the computer system, user input indicating user-specified concepts of interest, wherein the user input includes at least one keyword identified by the automated keyword analysis and at least one keyword not identified by the automated keyword analysis; analyzing, by the computer system, the electronic document to identify locations of discussion of the user-specified concepts of interest; and displaying, by the computer system, a graph representing the electronic document and illustrating persistence values associated with the user-specified concepts of interest at locations in the electronic document, wherein, for a given location in the electronic document, the graph illustrates a persistence value indicating a frequency of discussion of a user-specified concept of interest at that location relative to other locations in the electronic document. 4. The method of claim 1 wherein the graph comprises a first axis representing locations within the electronic document and a second axis representing persistence values of a user-specified concept of interest. | 0.761364 |
8,224,523 | 8 | 10 | 8. A vehicle communication system, in communication with a nomadic device, comprising: for a processor configured to receive a language or country designation as part of a packet sent from a communication point to the nomadic device, the designation having been added to the packet by the communication point; and wherein the processor is further configured to set a local-language emergency database (LLED) as a basis for a vehicle-spoken language when placing emergency calls, wherein, if an emergency call is originated by the vehicle computing system, outgoing spoken communication produced by the vehicle computing system is performed based on words and/or phrases stored in the LLED. | 8. A vehicle communication system, in communication with a nomadic device, comprising: for a processor configured to receive a language or country designation as part of a packet sent from a communication point to the nomadic device, the designation having been added to the packet by the communication point; and wherein the processor is further configured to set a local-language emergency database (LLED) as a basis for a vehicle-spoken language when placing emergency calls, wherein, if an emergency call is originated by the vehicle computing system, outgoing spoken communication produced by the vehicle computing system is performed based on words and/or phrases stored in the LLED. 10. The system of claim 8 , wherein the processor is further configured to establish communication between the vehicle computing system and the nomadic device, and a periodic signal to the nomadic device to transfer a language designation known by the nomadic device to the vehicle computing system. | 0.700401 |
9,626,447 | 11 | 16 | 11. A display device comprising: a display unit; a storage unit storing a table showing correspondences between text languages of web pages and character strings used in URLs to indicate the respective text languages; and a browser configured to acquire a web page from a web server over a network, and cause the display unit to display the web page acquired from the web server, wherein the browser comprises: a receiving unit receiving a designation of a URL; an acquiring unit acquiring information indicating a text language designated by a user; a first searching unit for searching for a top-level domain “com” or the top-level domain “com” with a slash “/” added to the end of the designated URL; a determining unit determining whether or not the designated URL includes, at the end thereof, the top-level domain “com” or the top-level domain “com” with a slash “/” added to the end thereof; a second searching unit, when the determining step determines that a top-level domain “com” or the top-level domain “com” with a slash “/” has been added to the end of the designated URL, acquiring source code of a web page indicated by the designated URL, and searching the acquired source code for a URL including a character string corresponding to the designated text language with reference to the table stored in the storage unit; and a display control unit, when the URL including the character string corresponding to the designated text language is found by the second searching unit, acquiring a web page indicated by the found URL from the web server over the network and displaying the acquired web page indicated by the found URL, and, when the URL including the character string corresponding to the designated text language is not found by the second searching unit, displaying the web page indicated by the designated URL according to the acquired source code. | 11. A display device comprising: a display unit; a storage unit storing a table showing correspondences between text languages of web pages and character strings used in URLs to indicate the respective text languages; and a browser configured to acquire a web page from a web server over a network, and cause the display unit to display the web page acquired from the web server, wherein the browser comprises: a receiving unit receiving a designation of a URL; an acquiring unit acquiring information indicating a text language designated by a user; a first searching unit for searching for a top-level domain “com” or the top-level domain “com” with a slash “/” added to the end of the designated URL; a determining unit determining whether or not the designated URL includes, at the end thereof, the top-level domain “com” or the top-level domain “com” with a slash “/” added to the end thereof; a second searching unit, when the determining step determines that a top-level domain “com” or the top-level domain “com” with a slash “/” has been added to the end of the designated URL, acquiring source code of a web page indicated by the designated URL, and searching the acquired source code for a URL including a character string corresponding to the designated text language with reference to the table stored in the storage unit; and a display control unit, when the URL including the character string corresponding to the designated text language is found by the second searching unit, acquiring a web page indicated by the found URL from the web server over the network and displaying the acquired web page indicated by the found URL, and, when the URL including the character string corresponding to the designated text language is not found by the second searching unit, displaying the web page indicated by the designated URL according to the acquired source code. 16. The display device according to claim 11 , wherein: the storage unit further stores therein a predetermined character string used in a URL of an index web page that includes a list of links to one or more web pages indicated by URLs including character strings indicating respective text languages, a browser program causes a computer to further perform: when the URL including the character string corresponding to the designated text language is not found in the second searching unit, determining whether or not the acquired source code includes the URL of the index web page by determining whether or not the acquired source code includes the URL including the predetermined character string; and when the acquired source code includes the URL including the predetermined character string, acquiring a source code of a web page indicated by the URL including the predetermined character string, and searching the acquired source code of the web page indicated by the URL including the predetermined character string for the URL including the character string corresponding to the designated text language, and when the URL including the character string corresponding to the designated text language is found in the second searching unit, the display control unit acquires a web page indicated by the URL found in the second searching unit from the web server over the network and displays the acquired web page indicated by the URL found in the second searching unit. | 0.5 |
8,010,669 | 11 | 13 | 11. An apparatus comprising a processor and memory including computer program code, the memory and computer program code configured to, with the processor, cause the apparatus to: receive property information from a provider node, the property information being descriptive of at least one of in-device property and at least one remote property having interface support; provide the property information to a delivery context client interface based context model; enable access, by a consumer application, to a property associated with the property information via the delivery context client interface based context model to enable provision of consumer data to the property to enable adaptive or customized services to the consumer application via the property; enable definition of an additional extension with respect to a property associated with the context model; and generate an event notification to indicate that a new property has been added including an additional extension to standard interfaces of the context model. | 11. An apparatus comprising a processor and memory including computer program code, the memory and computer program code configured to, with the processor, cause the apparatus to: receive property information from a provider node, the property information being descriptive of at least one of in-device property and at least one remote property having interface support; provide the property information to a delivery context client interface based context model; enable access, by a consumer application, to a property associated with the property information via the delivery context client interface based context model to enable provision of consumer data to the property to enable adaptive or customized services to the consumer application via the property; enable definition of an additional extension with respect to a property associated with the context model; and generate an event notification to indicate that a new property has been added including an additional extension to standard interfaces of the context model. 13. The apparatus of claim 11 , wherein the memory and computer program code are further configured to, with the processor, cause the apparatus to enable definition of the additional extension by providing a corresponding method identified by a method name, a parameter and an indication of a class of the parameter. | 0.524096 |
8,538,972 | 1 | 3 | 1. A method performed by a data processing apparatus, the method comprising: selecting, by the data processing apparatus, object representations from a dataset storing a plurality of object representations, each object representation being an association of: an object identifier that identifies an object instance in a dataset and corresponds to an object; a context value that identifies a context of the object; and a set of feature values that identify features of the object; wherein each object identifier is unique in the dataset, and each context value is associated with one or more object identifiers; for each feature value: determining an inter-context score that is proportional to the number of different context values in the dataset that are associated with the feature value, wherein the inter-context score is a probability that a random pair of object representations that each include a particular feature value are each associated with different context values and is determined independent of inter-context scores of other feature values; and determining an intra-context score that is proportional to the number of times the feature value is associated with each context value, wherein the intra-context score is a probability that a random pair of object representations that each include the particular feature value are each associated with the same context value and is determined independent of intra-context scores of other feature values; and for a selected pair of object representations, determining a similarity score based on inter-context scores and intra-context scores of matching feature values in the set of features for the pair of object representations, the similarity score being a measure of the similarity of the object representations in the pair of object representations, the determining comprising generating a respective set of feature weights for the set of features of the selected pair of object representations, each feature weight in the respective set of feature weights corresponding to a respective feature and being determined from the inter-context score and intra-context score for the respective feature and further determined independent of the inter-context score and intra-context score for other respective features. | 1. A method performed by a data processing apparatus, the method comprising: selecting, by the data processing apparatus, object representations from a dataset storing a plurality of object representations, each object representation being an association of: an object identifier that identifies an object instance in a dataset and corresponds to an object; a context value that identifies a context of the object; and a set of feature values that identify features of the object; wherein each object identifier is unique in the dataset, and each context value is associated with one or more object identifiers; for each feature value: determining an inter-context score that is proportional to the number of different context values in the dataset that are associated with the feature value, wherein the inter-context score is a probability that a random pair of object representations that each include a particular feature value are each associated with different context values and is determined independent of inter-context scores of other feature values; and determining an intra-context score that is proportional to the number of times the feature value is associated with each context value, wherein the intra-context score is a probability that a random pair of object representations that each include the particular feature value are each associated with the same context value and is determined independent of intra-context scores of other feature values; and for a selected pair of object representations, determining a similarity score based on inter-context scores and intra-context scores of matching feature values in the set of features for the pair of object representations, the similarity score being a measure of the similarity of the object representations in the pair of object representations, the determining comprising generating a respective set of feature weights for the set of features of the selected pair of object representations, each feature weight in the respective set of feature weights corresponding to a respective feature and being determined from the inter-context score and intra-context score for the respective feature and further determined independent of the inter-context score and intra-context score for other respective features. 3. The method of claim 1 , further comprising, for each set of feature values in the dataset, adding a unique feature value to the set of feature values and that is included only in that set of feature values, the unique feature value having an associated feature weight in a corresponding weighted vector that causes the cosine similarity value to be less than one. | 0.604752 |
8,898,202 | 15 | 20 | 15. A computer-readable storage device having instructions stored which, when executed by a computing device, cause the computing device to perform operations comprising: generating an input lattice representing input data having a first modality input and a second modality input; using a finite-state device to perform a function using the input lattice, the finite-state device having symbols based on markup-language semantics, wherein performing the function comprises mapping a first markup-language expression to a second markup-language expression using a transducer relating the first-markup-language expression to the second-markup-language expression based on a level of coincidence between the first modality input and the second modality input; generating a result finite-state device representing a result of the function, wherein the result finite-state device relates a set of input symbols to a set of output symbols; concatenating output symbols in the set of output symbols of the result finite-state device, to yield a markup-language expression; and outputting the markup-language expression. | 15. A computer-readable storage device having instructions stored which, when executed by a computing device, cause the computing device to perform operations comprising: generating an input lattice representing input data having a first modality input and a second modality input; using a finite-state device to perform a function using the input lattice, the finite-state device having symbols based on markup-language semantics, wherein performing the function comprises mapping a first markup-language expression to a second markup-language expression using a transducer relating the first-markup-language expression to the second-markup-language expression based on a level of coincidence between the first modality input and the second modality input; generating a result finite-state device representing a result of the function, wherein the result finite-state device relates a set of input symbols to a set of output symbols; concatenating output symbols in the set of output symbols of the result finite-state device, to yield a markup-language expression; and outputting the markup-language expression. 20. The computer-readable storage device of claim 15 , wherein the function performs multimodal integration of the input data. | 0.785714 |
8,165,878 | 1 | 12 | 1. A computer system method for matching an utterance of a user to a template comprising the steps of: (a) receiving by a processor the utterance from an input device, wherein the utterance includes at least one word; (b) accessing a set of template hierarchies from a database, wherein the set includes at least one template; (c) comparing by the processor the at least one word of the utterance to the at least one term of a template hierarchy in the set of template hierarchies; (d) determining by the processor whether the at least one word of the utterance matches the at least one term of the template hierarchy; (e) calculating by the processor a score based on the match between the at least one word of the utterance and the at least one term of the template hierarchy; (f) repeating steps (c)-(e) until there are no more words of the utterance for said comparing step; (g) populating the at least one template with at least one data element corresponding to the at least one term of the template hierarchy to obtain a populated template; (h) computing a total score based on the match between all words of the utterance to the populated template; (i) selecting by the processor the at least one template with the highest total score; (j) recording the populated template; and (k) communicating the populated template to the user. | 1. A computer system method for matching an utterance of a user to a template comprising the steps of: (a) receiving by a processor the utterance from an input device, wherein the utterance includes at least one word; (b) accessing a set of template hierarchies from a database, wherein the set includes at least one template; (c) comparing by the processor the at least one word of the utterance to the at least one term of a template hierarchy in the set of template hierarchies; (d) determining by the processor whether the at least one word of the utterance matches the at least one term of the template hierarchy; (e) calculating by the processor a score based on the match between the at least one word of the utterance and the at least one term of the template hierarchy; (f) repeating steps (c)-(e) until there are no more words of the utterance for said comparing step; (g) populating the at least one template with at least one data element corresponding to the at least one term of the template hierarchy to obtain a populated template; (h) computing a total score based on the match between all words of the utterance to the populated template; (i) selecting by the processor the at least one template with the highest total score; (j) recording the populated template; and (k) communicating the populated template to the user. 12. The computer system method of claim 1 wherein said calculating step further comprises the steps of: (i) representing a term in the template hierarchy as a term vector; (ii) obtaining an utterance term vector based on the match between a word of the utterance and a term of the template hierarchy; (iii) providing a score for the term of the template hierarchy based on the match between the term vector and utterance term vector; (iv) repeating steps (i)-(iii) until there are no more words of the utterance for said comparing step; and (v) selecting the term of the template hierarchy with the highest score. | 0.5 |
9,218,336 | 1 | 3 | 1. A method for constructing an automaton for automated analysis of an agglutinative language, the method comprising: constructing, using a processor of a computer, an affix automaton for each of a plurality of affix types of the agglutinative language, wherein each of said affix types is associated with one or more affixes associated with a morphological concept; combining, using the processor of the computer, any of said affix automatons to form a plurality of template automatons, where each of said template automatons is patterned after any of a plurality of agglutination templates of any of said affix types for said agglutinative language; combining, using the processor of the computer, said template automatons into a master automaton; receiving, by the processor of the computer, a word in the agglutinative language as an input for analysis; executing the master automaton to perform a morphological analysis of the received word, using the processor of the computer; and responsive to the executing, producing an output that indicates an expected part of speech for the word based on which of said template automatons were traversed within said master automaton during the executing. | 1. A method for constructing an automaton for automated analysis of an agglutinative language, the method comprising: constructing, using a processor of a computer, an affix automaton for each of a plurality of affix types of the agglutinative language, wherein each of said affix types is associated with one or more affixes associated with a morphological concept; combining, using the processor of the computer, any of said affix automatons to form a plurality of template automatons, where each of said template automatons is patterned after any of a plurality of agglutination templates of any of said affix types for said agglutinative language; combining, using the processor of the computer, said template automatons into a master automaton; receiving, by the processor of the computer, a word in the agglutinative language as an input for analysis; executing the master automaton to perform a morphological analysis of the received word, using the processor of the computer; and responsive to the executing, producing an output that indicates an expected part of speech for the word based on which of said template automatons were traversed within said master automaton during the executing. 3. The method according to claim 1 , wherein said constructing an affix automaton comprises constructing any of said affix automatons by starting with a last letter of each of said affixes and continuing towards a beginning letter of each of said affixes. | 0.785714 |
8,583,669 | 1 | 10 | 1. A method of suggesting a query for identifying documents in a litigation hold, comprising: receiving a training set of documents, wherein each document in the training set of documents is given a relevance indicator; identifying, by one or more processing devices, properties shared among documents having a similar relevance indicator, and an initial query with one or more initial keywords; determining a spatial proximity between one of the initial keywords and a set of keywords in the training set of documents; generating one or more neighboring queries from the initial query based on the spatial proximity, the neighboring queries including one or more additional queries; receiving a quality score of the neighboring queries, including an indication of a highest ranked neighboring query; repeating the generating one or more neighboring queries and receiving a quality score of the neighboring queries, wherein the highest ranked neighboring query is provided as the initial query, until an indication is received that a quality score of the highest ranked neighboring query is less than a quality score of the initial query, wherein a previously highest ranked neighboring query with a quality score greater than or equal to the quality score of the initial query is a resultant query; and generating a suggested litigation hold query based on the resultant query, wherein the suggested litigation hold query returns documents subject to the litigation hold. | 1. A method of suggesting a query for identifying documents in a litigation hold, comprising: receiving a training set of documents, wherein each document in the training set of documents is given a relevance indicator; identifying, by one or more processing devices, properties shared among documents having a similar relevance indicator, and an initial query with one or more initial keywords; determining a spatial proximity between one of the initial keywords and a set of keywords in the training set of documents; generating one or more neighboring queries from the initial query based on the spatial proximity, the neighboring queries including one or more additional queries; receiving a quality score of the neighboring queries, including an indication of a highest ranked neighboring query; repeating the generating one or more neighboring queries and receiving a quality score of the neighboring queries, wherein the highest ranked neighboring query is provided as the initial query, until an indication is received that a quality score of the highest ranked neighboring query is less than a quality score of the initial query, wherein a previously highest ranked neighboring query with a quality score greater than or equal to the quality score of the initial query is a resultant query; and generating a suggested litigation hold query based on the resultant query, wherein the suggested litigation hold query returns documents subject to the litigation hold. 10. The method of claim 1 , wherein the generating, one or more neighboring queries from the initial query comprises: selecting one or more additional keywords based on a random walk algorithm for use with one of the neighboring queries. | 0.762048 |
5,404,435 | 48 | 51 | 48. In a data processing system, a method for archiving voice objects in a document, comprising the steps of: loading an existing index into a data processing system; inputting a document architecture envelope including a text object and an voice object into said system; generating a first key word for said text object from said text object and adding said first key word to said index; automatically generating a second key word for said voice object from said text object and adding said second key word to said index; storing said document architecture envelope in said system; storing said index including said first and second key words in said system; entering a search term into said data processing system; comparing said search term with candidate key words in said index; and retrieving said voice object if said second key word is found in said comparing step. | 48. In a data processing system, a method for archiving voice objects in a document, comprising the steps of: loading an existing index into a data processing system; inputting a document architecture envelope including a text object and an voice object into said system; generating a first key word for said text object from said text object and adding said first key word to said index; automatically generating a second key word for said voice object from said text object and adding said second key word to said index; storing said document architecture envelope in said system; storing said index including said first and second key words in said system; entering a search term into said data processing system; comparing said search term with candidate key words in said index; and retrieving said voice object if said second key word is found in said comparing step. 51. The method of claim 48, wherein said second key word is generated from typing a word string into said system. | 0.536885 |
9,740,679 | 4 | 5 | 4. The method of claim 1 , wherein the method further comprises generating the reduced directed graph from an input directed graph by application of at least one reduction rule to the input directed graph, the input directed graph being generated from the n-gram statistics. | 4. The method of claim 1 , wherein the method further comprises generating the reduced directed graph from an input directed graph by application of at least one reduction rule to the input directed graph, the input directed graph being generated from the n-gram statistics. 5. The method of claim 4 , wherein the input directed graph includes a set of nodes connected by edges, each of the edges being labeled with a sequence of symbols and an associated multiplicity representing a number of occurrences of the sequence of symbols in the corpus sequence, each path through the input directed graph in which each edge is traversed the respective multiplicity of times reconstructing the corpus sequence from the labels in the order in which the edges are traversed. | 0.845888 |
8,909,534 | 1 | 15 | 1. A method comprising: selecting, by a computing device, sets of two or more text candidates from a plurality of text candidates corresponding to vocal input; for each set of the sets of two or more text candidates, outputting, by the computing device, representations of each of the respective two or more text candidates in the set to a plurality of users, wherein the representations are outputted as audio; receiving selections from each of the plurality of users of one of the respective two or more text candidates from each set, wherein the selections are based at least in part on satisfying a criterion; and determining that a text candidate included in the plurality of text candidates has a highest probability out of the plurality of text candidates of being a correct textual transcription of the vocal input based at least in part on the selections from the plurality of users, wherein the text candidate is a least-selected candidate by the plurality of users out of the plurality of text candidates. | 1. A method comprising: selecting, by a computing device, sets of two or more text candidates from a plurality of text candidates corresponding to vocal input; for each set of the sets of two or more text candidates, outputting, by the computing device, representations of each of the respective two or more text candidates in the set to a plurality of users, wherein the representations are outputted as audio; receiving selections from each of the plurality of users of one of the respective two or more text candidates from each set, wherein the selections are based at least in part on satisfying a criterion; and determining that a text candidate included in the plurality of text candidates has a highest probability out of the plurality of text candidates of being a correct textual transcription of the vocal input based at least in part on the selections from the plurality of users, wherein the text candidate is a least-selected candidate by the plurality of users out of the plurality of text candidates. 15. The method of claim 1 , wherein each set of two or more text candidates consists of two text candidates. | 0.864322 |
7,636,914 | 16 | 22 | 16. The medium of claim 13 , wherein said instructions further comprise: one or more instructions for evaluating said class at compile-time, and one or more instructions for adjusting said resulting class definitions from said evaluation to increase efficiency of run-time performance. | 16. The medium of claim 13 , wherein said instructions further comprise: one or more instructions for evaluating said class at compile-time, and one or more instructions for adjusting said resulting class definitions from said evaluation to increase efficiency of run-time performance. 22. The medium of claim 16 wherein said assignment function is used to identify and perform multiply and accumulate (“MAC”) operations. | 0.5 |
9,542,502 | 7 | 9 | 7. A data processing system comprising: a processor; and an accessible memory, the data processing system particularly configured to receive a document having fragments with attribute/value pairs; receive logical expressions that define relationships between fragments of the document; analyze the logical expressions to identify fragment names and attributes; create an index based on the analysis that includes names of the fragments to be candidates for selection into subdocuments; extract, from the document, all fragments named in the index; create, in the index, an entry for each attribute/value pair; create a plurality of subdocuments corresponding to the document, which includes finding top-level fragments in the index, finding second-level fragments related to each top-level fragment, and finding third-level fragments related to each second-level fragment; and store the subdocuments, including the respective related fragments. | 7. A data processing system comprising: a processor; and an accessible memory, the data processing system particularly configured to receive a document having fragments with attribute/value pairs; receive logical expressions that define relationships between fragments of the document; analyze the logical expressions to identify fragment names and attributes; create an index based on the analysis that includes names of the fragments to be candidates for selection into subdocuments; extract, from the document, all fragments named in the index; create, in the index, an entry for each attribute/value pair; create a plurality of subdocuments corresponding to the document, which includes finding top-level fragments in the index, finding second-level fragments related to each top-level fragment, and finding third-level fragments related to each second-level fragment; and store the subdocuments, including the respective related fragments. 9. The data processing system of claim 7 , wherein extracting the named fragments includes parsing values of each attribute according to the logical expressions. | 0.611111 |
6,012,158 | 3 | 4 | 3. A decoding apparatus for decoding an error correction code interleaved on a transmitting side and deinterleaved on a receiving side comprising: a determining circuit for making determination as to whether decoding is acceptable or not word by word according to a result of decoding; an error position detector for obtaining error position data by detecting an error position in decoding bit by bit according to the result of decoding; an estimator for estimating an error position according to error position data detected by said error position detector in a first word before a correction object word as well as to error position data detected by said error position detector in a second word after said correction object word when said determining circuit determines that decoding is not acceptable; a correcting circuit for correcting said correction object word according to the error position estimated by said estimator; and a redecoder for executing redecoding processing according to a result of correction by said correcting circuit. | 3. A decoding apparatus for decoding an error correction code interleaved on a transmitting side and deinterleaved on a receiving side comprising: a determining circuit for making determination as to whether decoding is acceptable or not word by word according to a result of decoding; an error position detector for obtaining error position data by detecting an error position in decoding bit by bit according to the result of decoding; an estimator for estimating an error position according to error position data detected by said error position detector in a first word before a correction object word as well as to error position data detected by said error position detector in a second word after said correction object word when said determining circuit determines that decoding is not acceptable; a correcting circuit for correcting said correction object word according to the error position estimated by said estimator; and a redecoder for executing redecoding processing according to a result of correction by said correcting circuit. 4. A decoding apparatus according to claim 3; wherein said error correction code is a BCH error correction code. | 0.692308 |
8,756,050 | 18 | 24 | 18. A computer-implemented method for providing translated content, comprising: providing, by one or more processors, to a first user a content instance in connection with an offer for consumption of an item in an electronic marketplace; receiving a translation of a content instance from the first user, the first user having a translator score, the translator score being based at least in part on particular translation scores for particular translations submitted by the first user and reviewer scores associated with one or more reviewers of the received translation; providing the received translation to the one or more reviewers; receiving one or more votes from a set of the one or more reviewers; calculating, based at least in part on the received one or more votes and the translator score, a translation score for the translation; and when the translation score satisfies one or more criteria, providing to a second user the translation of the content instance in connection with another offer for consumption of the item. | 18. A computer-implemented method for providing translated content, comprising: providing, by one or more processors, to a first user a content instance in connection with an offer for consumption of an item in an electronic marketplace; receiving a translation of a content instance from the first user, the first user having a translator score, the translator score being based at least in part on particular translation scores for particular translations submitted by the first user and reviewer scores associated with one or more reviewers of the received translation; providing the received translation to the one or more reviewers; receiving one or more votes from a set of the one or more reviewers; calculating, based at least in part on the received one or more votes and the translator score, a translation score for the translation; and when the translation score satisfies one or more criteria, providing to a second user the translation of the content instance in connection with another offer for consumption of the item. 24. The computer-implemented method of claim 18 , wherein a reviewer of the at least one reviewer of one or more reviewers, having a reviewer score above a threshold, impacts the translator score more heavily than a reviewer of the at least one reviewer of the one or more reviewers having reviewer scores below the threshold. | 0.601467 |
10,139,903 | 9 | 10 | 9. The system of claim 1 , wherein sequentially presenting successive components of the textual data comprises presenting a first word for a period of time, and, responsive to expiration of the period of time, implementing an animation transition before presenting a second word. | 9. The system of claim 1 , wherein sequentially presenting successive components of the textual data comprises presenting a first word for a period of time, and, responsive to expiration of the period of time, implementing an animation transition before presenting a second word. 10. The system of claim 9 , wherein the animation transition is a fading transition. | 0.511628 |
8,676,565 | 1 | 5 | 1. A method implemented by one or more computer processors, the method comprising: computationally matching one or more of a plurality of utterances into a respective one of a plurality of predefined topics through comparison with one or more corresponding semantic graph patterns: and for at least one said utterance that was not matched to the semantic graph patterns, computationally generating a semantic graph pattern that described information common to the at least one said utterance. | 1. A method implemented by one or more computer processors, the method comprising: computationally matching one or more of a plurality of utterances into a respective one of a plurality of predefined topics through comparison with one or more corresponding semantic graph patterns: and for at least one said utterance that was not matched to the semantic graph patterns, computationally generating a semantic graph pattern that described information common to the at least one said utterance. 5. A method as described in claim 1 , wherein the generating is based at least in part on conversation logs. | 0.668712 |
9,171,539 | 17 | 19 | 17. A method of converting components of a web page, having a plurality of HTML components, to voice prompts for a user, comprising: using an identifier and selecting a voice attribute file from a plurality of existing voice attribute files to be associated with a particular web page, the voice attribute file containing a plurality of voice attribute components that may be selectively enabled, the enabled voice attribute components in the selected voice attribute file determining which specific HTML components of the plurality of HTML components of a web page associated with the file are to be transformed for voice prompts; selecting a voice attribute component; determining whether the selected voice attribute component is enabled; in response to a determination that the selected voice attribute component is enabled, determining whether an HTML component of the plurality of HTML components is associated with the selected voice attribute component; determining whether a post-prompt associated with the HTML component and/or the selected voice attribute component is enabled; in response to a determination that a post-prompt associated with the HTML component and/or the selected voice attribute component is enabled, adding the post-prompt to a parameterized component; and forwarding the parameterized component to a mobile system, the mobile system configured to play a speech dialog using voice prompts based upon the parameterized component. | 17. A method of converting components of a web page, having a plurality of HTML components, to voice prompts for a user, comprising: using an identifier and selecting a voice attribute file from a plurality of existing voice attribute files to be associated with a particular web page, the voice attribute file containing a plurality of voice attribute components that may be selectively enabled, the enabled voice attribute components in the selected voice attribute file determining which specific HTML components of the plurality of HTML components of a web page associated with the file are to be transformed for voice prompts; selecting a voice attribute component; determining whether the selected voice attribute component is enabled; in response to a determination that the selected voice attribute component is enabled, determining whether an HTML component of the plurality of HTML components is associated with the selected voice attribute component; determining whether a post-prompt associated with the HTML component and/or the selected voice attribute component is enabled; in response to a determination that a post-prompt associated with the HTML component and/or the selected voice attribute component is enabled, adding the post-prompt to a parameterized component; and forwarding the parameterized component to a mobile system, the mobile system configured to play a speech dialog using voice prompts based upon the parameterized component. 19. The method of claim 17 , wherein the post-prompt prompts a user with a particular phrase. | 0.794248 |
4,718,094 | 73 | 74 | 73. The method of claim 71 comprising, the further step of: generating acoustic labels in response to the uttering of speech inputs wherein each label identifies a respective class of speech input, the classes being based on predefined features of speech input; and wherein the step of generating the first word score includes the step of: representing each vocabulary word by at least one probabilistic finite state phone machine, a representative phone machine having associated therewith (a) at least one transition, (b) a probability for each transition, and (c) actual label output probabilities each actual label output probability corresponding to the likelihood of the representative phone machine producing a given label at a given transition in response to an utterance, wherein each probability has a value assigned thereto. | 73. The method of claim 71 comprising, the further step of: generating acoustic labels in response to the uttering of speech inputs wherein each label identifies a respective class of speech input, the classes being based on predefined features of speech input; and wherein the step of generating the first word score includes the step of: representing each vocabulary word by at least one probabilistic finite state phone machine, a representative phone machine having associated therewith (a) at least one transition, (b) a probability for each transition, and (c) actual label output probabilities each actual label output probability corresponding to the likelihood of the representative phone machine producing a given label at a given transition in response to an utterance, wherein each probability has a value assigned thereto. 74. The method of claim 73 wherein the step of generating the first word score includes the further steps of: performing a detailed acoustic match for the subject word based on (i) the assigned values of the probabilities and (ii) the generated labels; and evaluating the first word score for the subject word from the performed detailed acoustic match. | 0.5 |
7,853,596 | 1 | 4 | 1. A method in a computing device for identifying a location associated with a first document, the method comprising: providing a collection of documents of words, each document labeled with an associated location, the collection not including the first document; generating by the computing device collection level parameters for a latent Dirichlet allocation style model for the collection of documents that is based on latent topics and the location of each document, the collection level parameters indicating a probability that a document in the collection relates to each latent topic, a probability that each word of the collection relates to each latent topic, and a probability that each location of the collection relates to each latent topic, wherein a variational expectation maximization algorithm is used to estimate the collection level parameters that are a maximization of a lower bound on the collection level parameters represented by a summation for each document in the collection of the log of the conditional probability of the document and its location given the collection level parameters; for each location, estimating, using the collection level parameters, a probability that the location is associated with the first document based on an aggregation of, for each topic, the conditional probability of the location given the topic and the conditional probability of the topic given the document, the conditional probabilities being derived from the collection of documents in which each document is labeled with an associated location; and selecting the location with the highest probability as the location associated with the first document. | 1. A method in a computing device for identifying a location associated with a first document, the method comprising: providing a collection of documents of words, each document labeled with an associated location, the collection not including the first document; generating by the computing device collection level parameters for a latent Dirichlet allocation style model for the collection of documents that is based on latent topics and the location of each document, the collection level parameters indicating a probability that a document in the collection relates to each latent topic, a probability that each word of the collection relates to each latent topic, and a probability that each location of the collection relates to each latent topic, wherein a variational expectation maximization algorithm is used to estimate the collection level parameters that are a maximization of a lower bound on the collection level parameters represented by a summation for each document in the collection of the log of the conditional probability of the document and its location given the collection level parameters; for each location, estimating, using the collection level parameters, a probability that the location is associated with the first document based on an aggregation of, for each topic, the conditional probability of the location given the topic and the conditional probability of the topic given the document, the conditional probabilities being derived from the collection of documents in which each document is labeled with an associated location; and selecting the location with the highest probability as the location associated with the first document. 4. The method of claim 1 including wherein the first document represents search terms of a query and wherein the probability for each location is estimated and the locations with the highest probability are indicated as locations that may be associated with the query. | 0.575949 |
6,079,047 | 8 | 12 | 8. In an integrated network holding first (14) and second (30) platform means each using respectively a first and second format, each said format being a different file format from the other and wherein said first platform means utilizes an original native data file of first format, and said second platform means (30) utilizes an industry standard byte stream data file of second format, a system for converting a second format Container (CD.backslash.FILE, 21d) of said second format files and directory from a CD-ROM (31d) into said original native data files in said first format for said first platform means, said system comprising: (a) means (31d, 33), in said second platform (30) means, for accessing said second format Container (CD.backslash.FILE,21d) holding said second format files and transferring said second format Container to a shared disk means (24); (b) said shared disk means (24) utilized by said first (14) and second (30) platform means for holding said second format Container (48) of said second format files (CD/FILE 48); (c) program means (36,38) for converting said second format Container (48) of second format files into separate original native first format files compatible for said first platform means. | 8. In an integrated network holding first (14) and second (30) platform means each using respectively a first and second format, each said format being a different file format from the other and wherein said first platform means utilizes an original native data file of first format, and said second platform means (30) utilizes an industry standard byte stream data file of second format, a system for converting a second format Container (CD.backslash.FILE, 21d) of said second format files and directory from a CD-ROM (31d) into said original native data files in said first format for said first platform means, said system comprising: (a) means (31d, 33), in said second platform (30) means, for accessing said second format Container (CD.backslash.FILE,21d) holding said second format files and transferring said second format Container to a shared disk means (24); (b) said shared disk means (24) utilized by said first (14) and second (30) platform means for holding said second format Container (48) of said second format files (CD/FILE 48); (c) program means (36,38) for converting said second format Container (48) of second format files into separate original native first format files compatible for said first platform means. 12. The system of claim 8 wherein said program means (36,38) includes: (c1) first interface program means (MCP.sub.-- FILEWRAPPER 36) to read out said second format Container (CD/FILE 48) on shared Disk C (24) for transfer to a second interface program (38); (c2) said second interface program (38, MCP.sub.-- WRAPPER) for converting said second format Container (CD/FILE 48) of second format files into a first format Container of first format files suitable for utilization by said first platform means. | 0.5 |
10,073,913 | 19 | 20 | 19. The server of claim 18 , wherein to generate a search query result set, the server is configured to transmit the search query to a search cluster and to receive ranked search result set therefrom, the search cluster having performed a general search. | 19. The server of claim 18 , wherein to generate a search query result set, the server is configured to transmit the search query to a search cluster and to receive ranked search result set therefrom, the search cluster having performed a general search. 20. The system of claim 19 , wherein to generate a search query result set, the server is further configured to transmit the search query to a plurality of vertical search modules and to receive vertical search results therefrom. | 0.5 |
9,280,525 | 15 | 18 | 15. A system for forming a structured document from an unstructured input document, the system comprising: a network connection that is configured to receive the input document from a data communication network, the input document being made available at a network address on the data communication network; one or more computers in electronic communication with the data communication network; a network address classifier present on various ones of the one or more computers and configuring the various ones of the one or more computers to determine whether data retrieved from the network address is possibly relevant to a plurality of textual classes; a translator present on various ones of the one or more computers and configuring the various ones of the one or more computers to extract a plurality of textual tokens from the input document, each extracted token having a visual style; a storage system for storing the extracted textual tokens and their visual styles; a content classifier present on various ones of the one or more computers and configuring the various ones of the one or more computers to produce, for each token, a first probability distribution of the given token with respect to the plurality of textual classes, the first probability distribution comprising a plurality of first probabilities each associated with a corresponding textual class of the plurality of textual classes such that each textual class has one of the plurality of first probabilities associated therewith, each of the plurality of textual classes being related to information conveyed by the textual tokens; a context classifier present on various ones of the one or more computers and configuring the various ones of the one or more computers to redistribute the first probability distribution of each token, based on the textual class having the highest first probability of the token's surrounding tokens in context, thereby producing a second probability distribution of the given token with respect to the plurality of textual classes, the second probability distribution comprising a plurality of second probabilities each associated with a corresponding textual class of the plurality of textual classes such that each textual class has one of the plurality of second probabilities associated therewith; and a visual style classifier present on various ones of the one or more computers and configuring the various ones of the one or more computers to produce, for each token based on its visual style and the second probability distribution, a third probability distribution of the given token with respect to the plurality of textual classes, the third probability distribution comprising a plurality of third probabilities each associated with a corresponding textual class of the plurality of textual classes such that each textual class has one of the plurality of third probabilities associated therewith; wherein each textual token is classified into one of the plurality of textual classes as a function of the second and third probability distributions. | 15. A system for forming a structured document from an unstructured input document, the system comprising: a network connection that is configured to receive the input document from a data communication network, the input document being made available at a network address on the data communication network; one or more computers in electronic communication with the data communication network; a network address classifier present on various ones of the one or more computers and configuring the various ones of the one or more computers to determine whether data retrieved from the network address is possibly relevant to a plurality of textual classes; a translator present on various ones of the one or more computers and configuring the various ones of the one or more computers to extract a plurality of textual tokens from the input document, each extracted token having a visual style; a storage system for storing the extracted textual tokens and their visual styles; a content classifier present on various ones of the one or more computers and configuring the various ones of the one or more computers to produce, for each token, a first probability distribution of the given token with respect to the plurality of textual classes, the first probability distribution comprising a plurality of first probabilities each associated with a corresponding textual class of the plurality of textual classes such that each textual class has one of the plurality of first probabilities associated therewith, each of the plurality of textual classes being related to information conveyed by the textual tokens; a context classifier present on various ones of the one or more computers and configuring the various ones of the one or more computers to redistribute the first probability distribution of each token, based on the textual class having the highest first probability of the token's surrounding tokens in context, thereby producing a second probability distribution of the given token with respect to the plurality of textual classes, the second probability distribution comprising a plurality of second probabilities each associated with a corresponding textual class of the plurality of textual classes such that each textual class has one of the plurality of second probabilities associated therewith; and a visual style classifier present on various ones of the one or more computers and configuring the various ones of the one or more computers to produce, for each token based on its visual style and the second probability distribution, a third probability distribution of the given token with respect to the plurality of textual classes, the third probability distribution comprising a plurality of third probabilities each associated with a corresponding textual class of the plurality of textual classes such that each textual class has one of the plurality of third probabilities associated therewith; wherein each textual token is classified into one of the plurality of textual classes as a function of the second and third probability distributions. 18. The system according to claim 15 , wherein the input document comprises an image, and the translator configures the various ones of the one or more computers to detect text in the image and extract the text as one or more of the plurality of textual tokens. | 0.5 |
8,259,124 | 1 | 2 | 1. A computer-implemented method for enhancing the visibility of keywords in search results, the method comprising: receiving search results from a search engine based on a search query, wherein the search query contains one or more keywords; identifying occurrences of keywords in the search results; displaying the search results in a user interface; applying a first highlight effect to identified occurrences of keywords in the displayed search results; detecting an occurrence of a highlighting change event; and after detecting the highlighting change event, animating a transition between the first highlight effect and a second highlight effect; and applying the second highlight effect to the displayed search results, such that a user receives an initial noticeable highlight that fades to a level that is still noticeable but is less likely to distract the user from other elements of the user interface. | 1. A computer-implemented method for enhancing the visibility of keywords in search results, the method comprising: receiving search results from a search engine based on a search query, wherein the search query contains one or more keywords; identifying occurrences of keywords in the search results; displaying the search results in a user interface; applying a first highlight effect to identified occurrences of keywords in the displayed search results; detecting an occurrence of a highlighting change event; and after detecting the highlighting change event, animating a transition between the first highlight effect and a second highlight effect; and applying the second highlight effect to the displayed search results, such that a user receives an initial noticeable highlight that fades to a level that is still noticeable but is less likely to distract the user from other elements of the user interface. 2. The method of claim 1 wherein receiving search results comprises receiving results asynchronously as the search engine finds matching results. | 0.599448 |
8,463,779 | 1 | 4 | 1. A method comprising: receiving, by at least one computing device, a request for one or more advertisements, each advertisement comprising content that is to be served for display in an initial web page; selecting, by the at least one computing device, one or more keyword candidates from the initial web page; performing, by the at least one computing device, a first query of a network for a first set of web pages containing a first keyword candidate from the one or more keyword candidates, the first query being performed using the first keyword candidate; identifying, by the at least one computing device, a contextual relevancy of the first keyword candidate to the initial web page by analyzing one or more of the web pages of the first set of web pages returned by the first query performed using the first keyword candidate; selecting, by the at least one computing device, one or more representative keywords from the one or more keyword candidates using the contextual relevancy of the first keyword candidate to the initial web page; selecting, by the at least one computing device in response to the request, the one or more advertisements using the one or more representative keywords selected using the identified contextual relevancy of the first keyword candidate to the initial web page; and transmitting, by the at least one computing device, the selected advertisement such that the selected advertisement is served for display with the initial web page. | 1. A method comprising: receiving, by at least one computing device, a request for one or more advertisements, each advertisement comprising content that is to be served for display in an initial web page; selecting, by the at least one computing device, one or more keyword candidates from the initial web page; performing, by the at least one computing device, a first query of a network for a first set of web pages containing a first keyword candidate from the one or more keyword candidates, the first query being performed using the first keyword candidate; identifying, by the at least one computing device, a contextual relevancy of the first keyword candidate to the initial web page by analyzing one or more of the web pages of the first set of web pages returned by the first query performed using the first keyword candidate; selecting, by the at least one computing device, one or more representative keywords from the one or more keyword candidates using the contextual relevancy of the first keyword candidate to the initial web page; selecting, by the at least one computing device in response to the request, the one or more advertisements using the one or more representative keywords selected using the identified contextual relevancy of the first keyword candidate to the initial web page; and transmitting, by the at least one computing device, the selected advertisement such that the selected advertisement is served for display with the initial web page. 4. The method of claim 1 wherein the one or more advertisements are served. | 0.918122 |
8,345,966 | 2 | 15 | 2. The method of claim 1 , further comprising computing a simplified representation of the scene corresponding to a human observation of the scene. | 2. The method of claim 1 , further comprising computing a simplified representation of the scene corresponding to a human observation of the scene. 15. The method of claim 2 , further comprising forming the vocabulary of color names, including: obtaining at least one general set of essential color names, each essential color name having a corresponding color value; wherein each color name includes a hue descriptor, brightness modifier, and saturation modifier; and selecting a subset of the at least one general set in meeting an application desire for the vocabulary of color names. | 0.71004 |
8,719,282 | 1 | 5 | 1. A computer-implemented method comprising: receiving a search query including a query term; receiving a document identified as being responsive to the search query; obtaining a synonym that a synonym rule identifies as a substitute for the query term in the search query; receiving an indication that the synonym of the query term in the search query is designated as a restricted-locality synonym; in response to receiving the indication that the synonym is designated as a restricted-locality synonym, selecting a first scoring model that specifies how to score occurrences of restricted-locality synonyms in documents, wherein the first scoring model is different than a second scoring model that specifies how to score occurrences of query terms or synonyms that are not designated as a restricted-locality synonym in documents, wherein the first scoring model specifies that a document score for the document depends on whether occurrences of the restricted-locality synonym in the document co-occur in the document with one or more query terms or one or more synonyms of the query term; and determining a document score for the document using the first scoring model. | 1. A computer-implemented method comprising: receiving a search query including a query term; receiving a document identified as being responsive to the search query; obtaining a synonym that a synonym rule identifies as a substitute for the query term in the search query; receiving an indication that the synonym of the query term in the search query is designated as a restricted-locality synonym; in response to receiving the indication that the synonym is designated as a restricted-locality synonym, selecting a first scoring model that specifies how to score occurrences of restricted-locality synonyms in documents, wherein the first scoring model is different than a second scoring model that specifies how to score occurrences of query terms or synonyms that are not designated as a restricted-locality synonym in documents, wherein the first scoring model specifies that a document score for the document depends on whether occurrences of the restricted-locality synonym in the document co-occur in the document with one or more query terms or one or more synonyms of the query term; and determining a document score for the document using the first scoring model. 5. The method of claim 1 , wherein the first scoring model uses one or more criteria for scoring occurrences of restricted-locality synonyms in the document. | 0.850476 |
8,345,934 | 2 | 3 | 2. The method according to claim 1 further comprising, if the number of character photos are less than n, selecting people photos for the subset of photos until reach n. | 2. The method according to claim 1 further comprising, if the number of character photos are less than n, selecting people photos for the subset of photos until reach n. 3. The method according to claim 2 further comprising, if the number of people photos are less than n, selecting non-people photos for the subset of photos according to their aesthetic measure until reach n. | 0.5 |
9,348,508 | 8 | 9 | 8. A computer program product for automatic detection of user preferences for an alternate user interface model, the computer program product comprising: a non-transitory computer readable storage medium for storing instructions for execution by a processing circuit for performing a method comprising: operating a digital device with an active user interface model, wherein the digital device includes a touchscreen that displays a user interface comprising a layout, and wherein the active user interface model is a first mapping of input gestures to a first set of operations to be executed by a processor of the digital device in response; receiving a repeated series of at least one input gesture from a user of the digital device via the touchscreen; determining a first likelihood ratio by comparing the series of input gestures with the first set of input gestures associated with the active user interface model; determining a second likelihood ratio by comparing the series of input gestures with a second set of input gesture associated with a latent user interface model, wherein the latent user interface model is a second mapping of input gestures to a second set of operations to be executed by the processor of the digital device in response, distinct from the first set of operations; determining the higher likelihood ratio from the first likelihood ratio and the second likelihood ratio; substituting the latent user interface model for the active user interface model, in response to the second likelihood ratio being higher, wherein substituting the latent user interface model does not alter the layout of the user interface; and in response to substituting the latent user interface model for the active user interface model, displaying a revert button on the touchscreen, wherein in response to a selection of the revert button the digital device reverts to the active user interface model. | 8. A computer program product for automatic detection of user preferences for an alternate user interface model, the computer program product comprising: a non-transitory computer readable storage medium for storing instructions for execution by a processing circuit for performing a method comprising: operating a digital device with an active user interface model, wherein the digital device includes a touchscreen that displays a user interface comprising a layout, and wherein the active user interface model is a first mapping of input gestures to a first set of operations to be executed by a processor of the digital device in response; receiving a repeated series of at least one input gesture from a user of the digital device via the touchscreen; determining a first likelihood ratio by comparing the series of input gestures with the first set of input gestures associated with the active user interface model; determining a second likelihood ratio by comparing the series of input gestures with a second set of input gesture associated with a latent user interface model, wherein the latent user interface model is a second mapping of input gestures to a second set of operations to be executed by the processor of the digital device in response, distinct from the first set of operations; determining the higher likelihood ratio from the first likelihood ratio and the second likelihood ratio; substituting the latent user interface model for the active user interface model, in response to the second likelihood ratio being higher, wherein substituting the latent user interface model does not alter the layout of the user interface; and in response to substituting the latent user interface model for the active user interface model, displaying a revert button on the touchscreen, wherein in response to a selection of the revert button the digital device reverts to the active user interface model. 9. The computer program product of claim 8 , wherein the latent user interface model is a plurality of latent user interface models, and the method further comprises, determining if one of the latent user interface models has a higher likelihood ratio than the first likelihood ratio of the active user interface models, wherein determining a likelihood ratio includes recognizing statistical properties of a sequence of user inputs that corresponds to one of the latent user interface models. | 0.5 |
9,959,069 | 17 | 18 | 17. A non-transitory computer-readable storage medium storing a text synchronization data structure relating to a first buffer, (1) textual user input having been applied to the first buffer, and (2) application text changes synchronized from a second buffer having been applied to the first buffer, the data structure comprising: one or more first entries, each first entry containing information identifying an editing action that (a) has been applied to the first buffer in response to textual user input, (b) has been communicated to the second buffer, and (c) has not been acknowledged by the second buffer, such that the contents of the first entries are usable to reverse editing actions that have been applied to the first buffer and are incompatible with application text changes to the second buffer. | 17. A non-transitory computer-readable storage medium storing a text synchronization data structure relating to a first buffer, (1) textual user input having been applied to the first buffer, and (2) application text changes synchronized from a second buffer having been applied to the first buffer, the data structure comprising: one or more first entries, each first entry containing information identifying an editing action that (a) has been applied to the first buffer in response to textual user input, (b) has been communicated to the second buffer, and (c) has not been acknowledged by the second buffer, such that the contents of the first entries are usable to reverse editing actions that have been applied to the first buffer and are incompatible with application text changes to the second buffer. 18. The non-transitory computer-readable storage medium of claim 17 , further comprising: one or more second entries, each second entry containing information identifying a character received as textual user input corresponding to an editing action identified by the information contained by a first entry, such that the content of the second entries are usable to replay to the first buffer the textual user input characters whose editing actions are incompatible with application text changes to the second buffer in response to reconciliation to the first buffer of the application text changes to the second buffer. | 0.5 |
7,711,569 | 1 | 7 | 1. A chat information system having a voice recognition device for recognizing voices, a voice synthesizer, a humanoid robot, a microphone for receiving the voices and a speaker for pronouncing synthesized voices, comprising: news capturing means for capturing news from the Internet; a news database for storing the captured news as a news reply, wherein the news reply has a plurality of priorities; a conversation database including at least a general conversation database storing a set of inquires and responses as conversation replies and a random chat database storing a plurality of predetermined responses, wherein a conversation reply has a plurality of priorities and at least one of the conversation replies belongs to one of a plurality of predetermined chat patterns; and a chat engine configured to: extract one or more keywords from a user's speech that has been recognized by the voice recognition device; determine whether a reserved chat flag associated with the extracted keywords is set, the reserved chat flag being set responsive to the extracted keywords belonging to one of the predetermined chat patterns; responsive to the reserved chat flag being set, setting a reply to a predetermined response of the plurality of the predetermined responses or a predetermined chat pattern of the plurality of the chat patterns; responsive to the reserved chat flag not being set, search at least one of the news database, the conversation database, and responsive to no matching being found from at least one of the news database and the conversation database, search the random chat database with the extracted keywords; select the reply among a plurality of replies found by the search based on the priorities of the replies; update the plurality of priorities associated with the replies; and output via the speaker the reply. | 1. A chat information system having a voice recognition device for recognizing voices, a voice synthesizer, a humanoid robot, a microphone for receiving the voices and a speaker for pronouncing synthesized voices, comprising: news capturing means for capturing news from the Internet; a news database for storing the captured news as a news reply, wherein the news reply has a plurality of priorities; a conversation database including at least a general conversation database storing a set of inquires and responses as conversation replies and a random chat database storing a plurality of predetermined responses, wherein a conversation reply has a plurality of priorities and at least one of the conversation replies belongs to one of a plurality of predetermined chat patterns; and a chat engine configured to: extract one or more keywords from a user's speech that has been recognized by the voice recognition device; determine whether a reserved chat flag associated with the extracted keywords is set, the reserved chat flag being set responsive to the extracted keywords belonging to one of the predetermined chat patterns; responsive to the reserved chat flag being set, setting a reply to a predetermined response of the plurality of the predetermined responses or a predetermined chat pattern of the plurality of the chat patterns; responsive to the reserved chat flag not being set, search at least one of the news database, the conversation database, and responsive to no matching being found from at least one of the news database and the conversation database, search the random chat database with the extracted keywords; select the reply among a plurality of replies found by the search based on the priorities of the replies; update the plurality of priorities associated with the replies; and output via the speaker the reply. 7. The system of claim 1 , wherein the reply is a captured news that has at least one or more words matching the extracted keywords of the user's speech. | 0.68125 |
9,361,280 | 14 | 15 | 14. The non-transitory computer storage medium of claim 11 , wherein maintaining the state information of the set of invoked activities includes maintaining state information for one or more background invoked activities, wherein the client-side code does not currently display the one or more background invoked activities on the client device. | 14. The non-transitory computer storage medium of claim 11 , wherein maintaining the state information of the set of invoked activities includes maintaining state information for one or more background invoked activities, wherein the client-side code does not currently display the one or more background invoked activities on the client device. 15. The non-transitory computer storage medium of claim 14 , wherein the state information of the set of invoked activities is maintained in an HTML DOM tree, wherein at least one of the plurality of live previews corresponds to a background invoked activity that is not currently displayed. | 0.5 |
8,321,463 | 14 | 16 | 14. One or more devices comprising: a comments database that stores comments associated with documents; a searching component to receive a request for comments associated with a particular document and to identify the comments by searching the comments database, where the request for comments is associated with a particular user; a ranking component to rank the identified comments based on a combination of objective scores and subjective scores, comprising: an objective ranking component to generate the objective scores for comments, where the objective scores are independent of the particular user, and a subjective ranking component to generate the subjective scores for comments, where the subjective scores depend on the particular user; and an interface to provide, to a client device associated with the user, the ranked comments for presentation in connection with the particular document. | 14. One or more devices comprising: a comments database that stores comments associated with documents; a searching component to receive a request for comments associated with a particular document and to identify the comments by searching the comments database, where the request for comments is associated with a particular user; a ranking component to rank the identified comments based on a combination of objective scores and subjective scores, comprising: an objective ranking component to generate the objective scores for comments, where the objective scores are independent of the particular user, and a subjective ranking component to generate the subjective scores for comments, where the subjective scores depend on the particular user; and an interface to provide, to a client device associated with the user, the ranked comments for presentation in connection with the particular document. 16. The one or more devices of claim 14 , where the objective ranking component generates the objective score for a particular comment based on objective ranking parameters that include at least one of: a measure of relevance of the particular comment to the particular document, a linking structure of the particular comment, a reputation of an author of the particular comment, a rating of the particular comment, a timestamp of the particular comment, a length of the particular comment, and a spelling and grammar score of the particular comment. | 0.5 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.