sentence1
stringlengths 16
446
| sentence2
stringlengths 14
436
|
|---|---|
in this paper , we proposed a new approach for analyzing the sentiment of figurative language . | therefore , the goal of our research is to find a new way to identify figurative meaning . |
in the context of arabic dialect translation , sawaf built a hybrid mt system that uses both statistical and rulebased approaches for da-to-english mt . | in the context of da translation , sawaf introduced a hybrid mt system that uses statistical and rule-based approaches for da-to-en mt . |
for training the model , we use the linear kernel svm implemented in the scikit-learn toolkit . | we use the selectfrommodel 4 feature selection method as implemented in scikit-learn . |
hoffmann et al use a probabilistic graphical model for multi-instance , multi-label learning and extract over newswire text using freebase relations . | hoffmann et al present a multi-instance multi-label model for relation extraction through distant supervision . |
we present novel evaluation paradigms for explanation methods for two classes of common nlp tasks ( see § 2 ) . | we have attempted to include all important local methods for nlp in our experiments ( see §3 ) . |
corpus pattern analysis is concerned with the prototypical syntagmatic patterns with which words in use are associated . | corpus pattern analysis attempts to catalog norms of usage for individual words , specifying them in terms of context patterns . |
we use the publicly available 300-dimensional word vectors of mikolov et al , trained on part of the google news dataset . | we use 300-dimensional vectors that were trained and provided by word2vec tool using a part of the google news dataset 4 . |
for our purpose we use word2vec embeddings trained on a google news dataset and find the pairwise cosine distances for all words . | first , we train a vector space representations of words using word2vec on chinese wikipedia . |
a residual connection is employed around each of two sub-layers , followed by layer normalization . | a residual connection and a layer normalization are then applied toq asq . |
to get a dictionary of word embeddings , we use the word2vec tool 2 and train it on the chinese gigaword corpus . | word representations to learn word embeddings from our unlabeled corpus , we use the gensim im-plementation of the word2vec algorithm . |
the language models in this experiment were trigram models with good-turing smoothing built using srilm . | we further used a 5-gram language model trained using the srilm toolkit with modified kneser-ney smoothing . |
figure 1 : a parse tree based on the treebank parse of wsj . | figure 7 : a parse produced by the unrestricted semantic model . |
xue et al proposed a translation-based language model for question retrieval . | xue et al proposed a word-based translation language model for question retrieval . |
in this paper , we propose an unsupervised approach for automatically detecting discussant . | in this paper , we presented an approach for subgroup detection in ideological discussions . |
the weights used during the reranking are tuned using the minimum error rate training algorithm . | their weights are optimized using minimum error-rate training on a held-out development set for each of the experiments . |
we use the 300-dimensional skip-gram word embeddings built on the google-news corpus . | we use the skip-gram model , trained to predict context tags for each word . |
in this paper , we propose to represent each word with an expressive multimodal distribution , for multiple distinct meanings . | in this paper , we propose , to the best of our knowledge , the first probabilistic word embedding that can capture multiple meanings . |
agrawal and an proposed a context-based approach to detect emotions from text at sentence level . | agrawal and an , 2012 ) proposed an unsupervised context-based approach to detect emotions from text at the sentence level . |
szarvas et al produced the bioscope corpus , which consists of biomedical texts annotated with negation and uncertainty , and their scopes . | szarvas et al present the bioscope corpus , which consists of medical and biological texts annotated for negation and speculation together with their linguistic scope . |
kalchbrenner et al showed that their dcnn for modeling sentences can achieve competitive results in this field . | kalchbrenner et al developed a cnnbased model that can be used for sentence modelling problems . |
we demonstrate the degree to which mt system rankings are dependent on weights employed in the construction of the gold standard . | we demonstrated the degree to which mt system rankings are dependent on weights employed in the construction of the gold standard . |
web-based models should therefore be used as a baseline for , rather than an alternative to , standard models . | rather , in our opinion , web-based models should be used as a new baseline for nlp tasks . |
our intuition is that there is a significant correlation between the sentiment of spoken text and an actually expressed emotion by the person . | the intuition is the same under m 4 , but now each token in a message is given its own class assignment , according to a class distribution for that particular message . |
we used the pre-trained word embeddings that are learned using the word2vec toolkit on google news dataset . | we used 300 dimensional skip-gram word embeddings pre-trained on pubmed . |
a tri-gram language model is estimated using the srilm toolkit . | a standard sri 5-gram language model is estimated from monolingual data . |
we trained the five classifiers using the svm implementation in scikit-learn . | we used the svm implementation provided within scikit-learn . |
we tokenised and parsed the text to obtain dependency trees , using the stanford parser . | we used the stanford neural network parser to obtain dependency triples . |
based on , rockt盲schel et al uses the attention-based technique to improve the performance of lstm-based recurrent neural network . | rockt盲schel et al propose neural network with attention mechanism , making neural networks interpretable . |
relation extraction is the task of predicting semantic relations over entities expressed in structured or semi-structured text . | relation extraction is a core task in information extraction and natural language understanding . |
to minimize the objective , we use the diagonal variant of adagrad with minibatches . | we use the diagonal variant of adagrad with minibatches , which is widely applied in deep learning literature , . |
yessenalina and cardie represent each word as a matrix and use iterated matrix multiplication as phrase-level composition function . | yessenalina and cardie modeled each word as a matrix and used iterated matrix multiplication to present a phrase . |
in this work , we proposed three new methods for training neural network language models and showed their efficiency both in terms of computational complexity and generalization performance . | in this work , we study the performance and behavior of two neural statistical language models so as to highlight some important caveats of the classical training algorithms . |
the data comes from the conll 2000 shared task , which consists of sentences from the penn treebank wall street journal corpus . | the data consist of four-tuples of words , extracted from the wall street journal treebank by a group at ibm . |
corpus-derived models of semantics have been extensively studied in the nlp and machine learning communities . | vector-space models of lexical semantics have been a popular and effective approach to learning representations of word meaning . |
we set the feature weights by optimizing the bleu score directly using minimum error rate training on the development set . | we tune weights by minimizing bleu loss on the dev set through mert and report bleu scores on the test set . |
the srilm toolkit was used to build this language model . | the srilm toolkit is used to train 5-gram language model . |
the benchmark model for topic modelling is latent dirichlet allocation , a latent variable model of documents . | latent dirichlet allocation is one of the widely adopted generative models for topic modeling . |
in recent years , various phrase translation approaches have been shown to outperform word-to-word translation models . | phrase-based statistical machine translation models have achieved significant improvements in translation accuracy over the original ibm word-based model . |
we used a phrase-based smt model as implemented in the moses toolkit . | we implemented our method in a phrase-based smt system . |
neelakantan et al proposed an extension of the skip-gram model combined with context clustering to estimate the number of senses for each word as well as learn sense embedding vectors . | neelakantan et al proposed the multisense skip-gram model , that jointly learns context cluster prototypes and word sense embeddings . |
in a relatively high-dimensional feature space may suffer from the data sparseness problem . | however , the richer feature representations result in a high-dimensional feature space . |
to calculate language model features , we train traditional n-gram language models with ngram lengths of four and five using the srilm toolkit . | we then lowercase all data and use all sentences from the modern dutch part of the corpus to train an n-gram language model with the srilm toolkit . |
however , in practice , there are many domains , such as the biomedical domain , which involve nested , overlapping , discontinuous ne mentions . | however , in practice , there are many domains , such as the biomedical domain , in which there are nested , overlapping , and discontinuous entity mentions . |
some recent work on active learning has started to include more realistic measures of the actual costs of annotation . | however , recently there is increased interest in measuring the true costs of annotation work when doing active learning . |
we apply this approach to a knowledge base of approximately 500 , 000 beliefs extracted imperfectly from the web by nell . | we applied this approach to a knowledge base of approximately 500,000 beliefs extracted imperfectly from the web by nell . |
in this paper , we present finite structure query ( fsq ) , a query tool for syntactically annotated . | in this paper , we presented fsq , a query tool for syntactically annotated corpora . |
we present the inesc-id system for the 2015 semeval message polarity classification task . | we have presented the inesc-id system for the semeval 2015 message classification task . |
topic assignment of each word is not independent , but rather affected by the topic . | the topic assignment for each word is irrelevant to all other words . |
we train a cnn with one layer of convolution and max pooling on top of word embedding vectors trained on the google news corpus of size 300 . | we use word embedding vectors trained on the google news corpus of size 300 , to train a cnn with one layer of convolution and max pooling . |
coreference resolution is the task of determining whether two or more noun phrases refer to the same entity in a text . | coreference resolution is the process of linking together multiple referring expressions of a given entity in the world . |
we used the scikit-learn implementation of a logistic regression model using the default parameters . | within this subpart of our ensemble model , we used a svm model from the scikit-learn library . |
sentiment analysis ( cite-p-12-3-17 ) is a popular research topic which has a wide range of applications , such as summarizing customer reviews , monitoring social media , and predicting stock market trends ( cite-p-12-1-4 ) . | sentiment analysis ( cite-p-8-1-20 ) is a task of predicting whether the text expresses a positive , negative , or neutral opinion in general or with respect to an entity of interest . |
a bunsetsu consists of one independent word and more than zero ancillary words . | 1 a bunsetsu is the linguistic unit in japanese that roughly corresponds to a basic phrase in english . |
in this paper , we conduct a detailed study of the causes of spurious ambiguity . | however , to our knowledge , we give the first detailed analysis on spurious ambiguity of word alignment . |
we use pre-trained glove vector for initialization of word embeddings . | for the actioneffect embedding model , we use pre-trained glove word embeddings as input to the lstm . |
we pre-trained word embeddings using word2vec over tweet text of the full training data . | we trained word vectors with the two architectures included in the word2vec software . |
semantic textual similarity is the task of measuring the degree to which two texts have the same meaning . | semantic textual similarity is the task of measuring the degree to which two text snippets have the same meaning . |
coreference resolution is a key problem in natural language understanding that still escapes reliable solutions . | coreference resolution is the task of clustering a sequence of textual entity mentions into a set of maximal non-overlapping clusters , such that mentions in a cluster refer to the same discourse entity . |
this study encodes distributional semantics into the triple-based background knowledge ranking model for better document enrichment . | zhang et al proposed a triple-based document enrichment framework which uses triples of spo as background knowledge . |
named entity recognition ( ner ) is the task of detecting named entity mentions in text and assigning them to their corresponding type . | named entity recognition ( ner ) is a frequently needed technology in nlp applications . |
word sense disambiguation ( wsd ) is a particular problem of computational linguistics which consists in determining the correct sense for a given ambiguous word . | word sense disambiguation ( wsd ) is the task of determining the correct meaning for an ambiguous word from its context . |
the system is based on a statistical model whose parameters are trained discriminatively using annotated sentences in the amr bank corpus . | the approach is a statistical natural language generation system , trained discriminatively using sentences in the amr bank . |
in the task-6 results ( cite-p-15-1-4 ) , our system was ranked 21th out of 85 participants with 0 . 6663 pearson-correlation . | in the task-6 results ( cite-p-15-1-4 ) , our system was ranked 21th out of 85 participants with 0.6663 pearson-correlation all competition rank . |
solve this problem , we often first read each piece of text , collect some answer candidates , then focus on these candidates and combine their information to select the final answer . | we first extract answer candidates from passages , then select the final answer by combining information from all the candidates . |
chang and han , sun and xu used rich statistical information as discrete features in a sequence labeling framework . | sun and xu enhanced the segmentation results by interpolating the statistics-based features derived from unlabeled data to a crfs model . |
a handful of papers have leveraged this idea for summarization . | a handful of papers have studied system combination for summarization . |
tuning is performed to maximize bleu score using minimum error rate training . | the decoding weights are optimized with minimum error rate training to maximize bleu scores . |
the feature weights are tuned to optimize bleu using the minimum error rate training algorithm . | the log-linear model is then tuned as usual with minimum error rate training on a separate development set coming from the same domain . |
it has been shown that images from google yield higher quality representations than comparable resources such as flickr and are competitive with hand-crafted datasets . | it has been shown that images from google yield higher-quality representations than comparable sources such as flickr . |
dependency parsing is a valuable form of syntactic processing for nlp applications due to its transparent lexicalized representation and robustness with respect to flexible word order languages . | dependency parsing is a basic technology for processing japanese and has been the subject of much research . |
we used pos tags predicted by the stanford pos tagger . | we depend on stanford pos tagger for getting pos tags of the corpus . |
language models were built using the sri language modeling toolkit with modified kneser-ney smoothing . | language models were built with srilm , modified kneser-ney smoothing , default pruning , and order 5 . |
we then used word2vec to train word embeddings with 512 dimensions on each of the prepared corpora . | the word embeddings required by our proposed methods were trained using the gensim 5 implementation of the skip gram version of word2vec . |
target language models were trained on the english side of the training corpus using the srilm toolkit . | additionally , a back-off 2-gram model with goodturing discounting and no lexical classes was built from the same training data , using the srilm toolkit . |
mikolov et al proposed a distributed word embedding model that allowed to convey meaningful information on vectors derived from neural networks . | mikolov et al presents a neural network-based architecture which learns a word representation by learning to predict its context words . |
in addition to improving the original k & m noisy-channel model , we create unsupervised and semi-supervised models of the task . | we have created a supervised version of the noisy-channel model with some improvements over the k & m model . |
our experiments indicate that mem significantly outperforms prior work in both sentence-level rating . | our experiments indicate that mem achieves better overall accuracy than alternative methods . |
mikolov et al proposed a computationally efficient method for learning distributed word representation such that words with similar meanings will map to similar vectors . | mikolov et al proposed the word2vec method for learning continuous vector representations of words from large text datasets . |
computational linguistics , volume 14 , number 3 , september 1988 47 quilici , dyer , and flowers recognizing and responding to plan-oriented misconceptions . | 44 computational linguistics , volume 14 , number 3 , september 1988 quilici , dyer , and flowers recognizing and responding to plan-oriented misconceptions |
with ca . 1000 instances , the proposed method increases the macro-average f-score and accuracy up to 50 % , compared to a baseline classifier . | for instance , with training sets of c.a . 1000 labeled instances , the proposed method brings improvements in accuracy and macro-average f-score up to 50 % compared to a baseline classifier . |
using latent topical dimensions , the model is able to discriminate between different senses . | using the latent space , the model is able to discriminate between different word senses . |
a typical user can most readily supply and identify the tables . | users typically know the database structure and contents . |
we describe our contribution to the semeval-2015 shared task : sentiment analysis of figurative language in twitter . | this paper describes our contribution to the semeval-2015 task 11 on sentiment analysis of figurative language in twitter . |
another approach is taken by , where , based on source and target language models , the authors calculated the difference of the cross-entropy values for a given sentence . | moore and lewis calculated the difference of the cross entropy values for a given sentence , based on language models from the source domain and the target domain . |
galley and manning introduce the hierarchical phrase reordering model which increases the consistency of orientation assignments . | for standard phrase-based translation , galley and manning introduced a hierarchical phrase orientation model . |
we trained a 5-gram language model on the xinhua portion of gigaword corpus using the srilm toolkit . | a 4-gram language model is trained on the xinhua portion of the gigaword corpus with the srilm toolkit . |
we build an open-vocabulary language model with kneser-ney smoothing using the srilm toolkit . | we use the sri language model toolkit to train a 5-gram model with modified kneser-ney smoothing on the target-side training corpus . |
bagga and baldwin , 1998b ) presented one of the first cdc systems , which relied solely on the contextual words of the named entities . | bagga and baldwin , 1998 ) proposed a method using the vector space model to disambiguate references to a person , place , or event across multiple documents . |
relation extraction is the key component for building relation knowledge graphs , and it is of crucial significance to natural language processing applications such as structured search , sentiment analysis , question answering , and summarization . | relation extraction ( re ) is the task of recognizing the assertion of a particular relationship between two or more entities in text . |
that , due to its computational complexity , it is difficult to straightforwardly apply previously studied techniques of bilingual term correspondence estimation from comparable corpora , especially in the case of large scale evaluation such as those presented in this paper . | first , we show that , due to its computational complexity , it is difficult to straightforwardly apply previously studied techniques of bilingual term correspondence estimation from comparable corpora , especially in the case of large scale evaluation such as those presented in this paper . |
semantic role labeling ( srl ) is the task of labeling predicate-argument structure in sentences with shallow semantic information . | semantic role labeling ( srl ) is the task of identifying semantic arguments of predicates in text . |
for the textual sources , we populate word embeddings from the google word2vec embeddings trained on roughly 100 billion words from google news . | we train skip-gram word embeddings with the word2vec toolkit 1 on a large amount of twitter text data . |
we then lowercase all data and use all unique headlines in the training data to train a language model with the srilm toolkit . | we use srilm for training a trigram language model on the english side of the training data . |
speech is a major component of modern user interfaces as it is the natural means of human communication . | speech is a single step within a larger system . |
bilingual dictionaries are an essential resource in many multilingual natural language processing tasks such as machine translation and cross-language information retrieval . | bilingual lexicons play an important role in many natural language processing tasks , such as machine translation and cross-language information retrieval . |
an alternative approach is based on a continuous representation of the words . | this lm approach is based a continuous representation of the words . |
and , thus , reflects a better lexical choice of the content words . | furthermore , the models are often capable to produce a better lexical choice of content words . |
semantic role labeling ( srl ) is the task of automatic recognition of individual predicates together with their major roles ( e.g . frame elements ) as they are grammatically realized in input sentences . | semantic role labeling ( srl ) is a kind of shallow sentence-level semantic analysis and is becoming a hot task in natural language processing . |
the third baseline , a bigram language model , was constructed by training a 2-gram language model from the large english ukwac web corpus using the srilm toolkit with default good-turing smoothing . | an n-gram language model was then built from the sinica corpus released by the association for computational linguistics and chinese language processing using the srilm toolkit . |
our baseline is a phrase-based mt system trained using the moses toolkit . | our smt system is a phrase-based system based on the moses smt toolkit . |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.