Datasets:
lguenth
/
backsum

Dataset card Data Studio Files Community
1
source_text
stringlengths
27
368
label
int64
0
1
target_text
stringlengths
1
5.38k
While many systems had similar performance, the results offered interesting insights, especially, about the relative performance of statistical and rule-based systems.
0
Following this method, we repeatedly — say, 1000 times — sample sets of sentences from the output of each system, measure their BLEU score, and use these 1000 BLEU scores as basis for estimating a confidence interval.
The TRIE model uses less memory than the smallest lossless alternative and is still faster than SRILM.
0
MITLM 0.4 (Hsu and Glass, 2008) is mostly designed for accurate model estimation, but can also compute perplexity.
The authors in this paper describe a search procedure for statistical machine translation (MT) based on dynamic programming (DP).
0
German English Training: Sentences 58 073 Words 519 523 549 921 Words* 418 979 453 632 Vocabulary Size 7939 4648 Singletons 3454 1699 Test-147: Sentences 147 Words 1 968 2 173 Perplexity { 26:5 Table 4: Multi-reference word error rate (mWER) and subjective sentence error rate (SSER) for three different search procedures.
In this paper the authors present a stochastic finite-state model for segmenting Chinese text into words.
0
In addition to the automatic methods, AG, GR, and ST, just discussed, we also added to the plot the values for the current algorithm using only dictionary entries (i.e., no productively derived words or names).
In this paper the author evaluates machine translation performance for six European language pairs that participated in a shared task: translating French, German, Spanish texts to English and back.
0
At the very least, we are creating a data resource (the manual annotations) that may the basis of future research in evaluation metrics.
There are clustering approaches that assign a single POS tag to each word type.
0
The relevant variables are the set of token-level tags that appear before and after each instance of the ith word type; we denote these context pairs with the set {(tb, ta)} and they are contained in t(−i).
Koehn and Monz carried out an extensive manual and automatic evaluation of machine translation performance on European language pairs.
0
They demonstrated this with the comparison of statistical systems against (a) manually post-edited MT output, and (b) a rule-based commercial system.
This paper talks about KenLM: Faster and Smaller Language Model Queries.
0
Unigram records store probability, backoff, and an index in the bigram table.
The resulting model is compact, efficiently learnable and linguistically expressive.
0
7 Conclusion and Future Work.
The authors in this paper describe a search procedure for statistical machine translation (MT) based on dynamic programming (DP).
0
The following two error criteria are used in our experiments: mWER: multi-reference WER: We use the Levenshtein distance between the automatic translation and several reference translations as a measure of the translation errors.
The manual evaluation of scoring translation on a graded scale from 1–5 seems to be very hard to perform.
0
We settled on contrastive evaluations of 5 system outputs for a single test sentence.
The use of global features has shown excellent result in the performance on MUC-6 and MUC-7 test data.
0
Since MUC6, BBN' s Hidden Markov Model (HMM) based IdentiFinder (Bikel et al., 1997) has achieved remarkably good performance.
This paper talks about KenLM: Faster and Smaller Language Model Queries.
0
Process statistics are already collected by the kernel (and printing them has no meaningful impact on performance).
The approach has been successfully tested on the 8 000-word Verbmobil task.
0
Source sentence words are aligned with hypothesized target sentence words, where the choice of a new source word, which has not been aligned with a target word yet, is restricted1.
The evaluation compares the performance of the system with that of several human judges and inter-human agreement on a single correct way to segment a text.
0
On a set of 11 sentence fragments-the A set-where they reported 100% recall and precision for name identification, we had 73% recall and 80% precision.
A beam search concept is applied as in speech recognition.
0
Although the ultimate goal of the Verbmobil project is the translation of spoken language, the input used for the translation experiments reported on in this paper is the (more or less) correct orthographic transcription of the spoken sentences.
NER is useful in many NLP applications such as information extraction, question answering, etc .Chieu and Ng have shown that the maximum entropy framework is able to use global information directly from various sources.
0
If they are found in a list, then a feature for that list will be set to 1.
The model incorporates various recent techniques for incorporating and manipulating linguistic knowledge using finite-state transducers.
0
This is a rather important source of errors in name identifi­ cation, and it is not really possible to objectively evaluate a name recognition system without considering the main lexicon with which it is used.
A beam search concept is applied as in speech recognition.
0
(f1; ;mg ; l) 2 (f1; ;mg n fl; l1g ; l0) !
Throughout this paper we compare with several packages: SRILM 1.5.12 (Stolcke, 2002) is a popular toolkit based on tries used in several decoders.
0
Syntactic decoders, such as cdec (Dyer et al., 2010), build state from null context then store it in the hypergraph node for later extension.
They believe that global context is useful in most languages, as it is a natural tendency for authors to use abbreviations on entities already mentioned previously.
0
The local feature groups are: Non-Contextual Feature: This feature is set to 1 for all tokens.
The overall parsing accuracy obtained with the pseudo-projective approach is still lower than for the best projective parsers.
0
This work was supported in part by the Swedish Research Council (621-2002-4207).
The authors in this paper describe a search procedure for statistical machine translation (MT) based on dynamic programming (DP).
0
German city names are replaced by category markers.
This paper talks about Exploiting Diversity in Natural Language Processing: Combining Parsers.
0
All of these systems were run on data that was not seen during their development.
The first method builds on results from (Yarowsky 95) and (Blum and Mitchell 98).
0
The Expectation Maximization (EM) algorithm (Dempster, Laird and Rubin 77) is a common approach for unsupervised training; in this section we describe its application to the named entity problem.
The PROBING data structure uses linear probing hash tables and is designed for speed.
0
The PROBING model can perform optimistic searches by jumping to any n-gram without needing state and without any additional memory.
The approach assumes that the word reordering is restricted to a few positions in the source sentence.
0
The proof is given in (Tillmann, 2000).
They showed better grammars to improve performance on both morphological and syntactic tasks, providing support for the advantage of a joint framework over pipelined or factorized ones.
0
In turn we use two sorts of heuristics, orthogonal to one another, to prune segmentation possibilities based on lexical and grammatical constraints.
Their results suggested that it was possible to learn accurate POS taggers for languages which did not have any annotated data, but have translations into a resource-rich language.
0
As a result, its POS tag needs to be induced in the “No LP” case, while the 11A word level paired-t-test is significant at p < 0.01 for Danish, Greek, Italian, Portuguese, Spanish and Swedish, and p < 0.05 for Dutch. correct tag is available as a constraint feature in the “With LP” case.
They used graph-based label propagation for cross-lingual knowledge transfer and used the projected labels as features in an unsupervised model.
0
In general, the neighborhoods can be more diverse and we allow a soft label distribution over the vertices.
The corpus was annoted with different linguitic information.
0
The corpus has been annotated with six different types of information, which are characterized in the following subsections.
BABAR's performance in both domains of terrorism and natural disaster, and the contextual-role knowledge in pronouns have shown successful results.
0
The inventions disclosed herein are the subject of a patent application owned by the University of Utah and licensed on an exclusive basis to Attensity Corporation.
Their results suggested that it was possible to learn accurate POS taggers for languages which did not have any annotated data, but have translations into a resource-rich language.
0
They considered a semi-supervised POS tagging scenario and showed that one can use a graph over trigram types, and edge weights based on distributional similarity, to improve a supervised conditional random field tagger.
This paper offers a broad insight into of Arabic constituency parsing by analyzing the interplay of linguistic phenomena, annotation choices, and model design.
0
Ends with the feminine affix :: p. 4.
In this paper, the authors are of the opinion that the sequence models-based approaches usually treat token-level tag assignment as the primary latent variable.
0
β is the shared hyperparameter for the tag assignment prior and word feature multinomials.
This paper conducted research in the area of automatic paraphrase discovery.
0
In the “Person – Person” domain, 618 keywords are found, and in the “Country – Country” domain, 303 keywords are found.
The overall parsing accuracy obtained with the pseudo-projective approach is still lower than for the best projective parsers.
0
However, it can be noted that the results for the least informative encoding, Path, are almost comparable, while the third encoding, Head, gives substantially worse results for both data sets.
The authors show that PATB is similar to other tree-banks but that annotation consistency remains low.
0
The ATB gives several different analyses to these words to indicate different types of coordination.
Vijay-Shankar et all considered the structural descriptions produced by various grammatical formalisms in terms of the complexity of the paths and the relationship between paths in the sets of structural descriptions that each system can generate.
0
Thus, the tree sets generated by HG's are similar to those of CFG's, with each node annotated by the operation (concatenation or wrapping) used to combine the headed strings derived by the daughters of Tree Adjoining Grammars, a tree rewriting formalism, was introduced by Joshi, Levy and Takahashi (1975) and Joshi (1983/85).
The bias of automatic methods in favour of statistical systems seems to be less pronounced on out-of-domain test data.
0
The development of automatic scoring methods is an open field of research.
The authors use linguistic and annotation insights to develop a manually annotated grammar and evaluate it and finally provide a realistic evaluation in which segmentation is performed in a pipeline jointly with parsing.
0
evaluated to account for the same fraction of the data.
Nevertheless, only a part of this corpus (10 texts), which the authors name "core corpus", is annotated with all this information.
0
Due to the dramatic fiscal situation in Brandenburg she now surprisingly withdrew legislation drafted more than a year ago, and suggested to decide on it not before 2003.
BABAR's performance in both domains of terrorism and natural disaster, and the contextual-role knowledge in pronouns have shown successful results.
0
BABAR uses a named entity recognizer to identify proper names that refer to people and companies.
they evaluate various specific aspects of the segmentation, as well as the overall segmentation performance.
0
A related point is that mutual information is helpful in augmenting existing electronic dictionaries, (cf.
Finally, several coreference systems have successfully incorporated anaphoricity determination modules.
0
This work was supported in part by the National Science Foundation under grant IRI9704240.
It is well-known that English constituency parsing models do not generalize to other languages and treebanks.
0
How should the absence of vowels and syntactic markers influence annotation choices and grammar development?
This paper presents methods to query N-gram language models, minimizing time and space costs.
0
Floating point values may be stored in the trie exactly, using 31 bits for non-positive log probability and 32 bits for backoff5.
Vijay-Shankar et all considered the structural descriptions produced by various grammatical formalisms in terms of the complexity of the paths and the relationship between paths in the sets of structural descriptions that each system can generate.
0
This suggests that by generalizing the notion of context-freeness in CFG's, we can define a class of grammatical formalisms that manipulate more complex structures.
It is annotated with several data: morphology, syntax, rhetorical structure, connectors, correference and informative structure.
0
This paper, however, provides a comprehensive overview of the data collection effort and its current state.
Their method did not assume any knowledge about the target language, making it applicable to a wide array of resource-poor languages.
0
They considered a semi-supervised POS tagging scenario and showed that one can use a graph over trigram types, and edge weights based on distributional similarity, to improve a supervised conditional random field tagger.
It also incorporates the Good-Turing methodin estimating the likelihoods of previously unseen constructions, including morphological derivatives and personal names.
0
An examination of the subjects' bracketings confirmed that these instructions were satisfactory in yielding plausible word-sized units.
NER is useful in many NLP applications such as information extraction, question answering, etc .Chieu and Ng have shown that the maximum entropy framework is able to use global information directly from various sources.
0
Hence, for each token, one of the four features zone-TXT, zone- HL, zone-DATELINE, or zone-DD is set to 1, and the other 3 are set to 0.
They plan on extending instance-weighting to other standard SMT components and capture the degree of generality of phrase pairs.
0
This is motivated by taking β po(s|t) to be the parameters of a Dirichlet prior on phrase probabilities, then maximizing posterior estimates p(s|t) given the IN corpus.
An extended lexicon model is defined, and its likelihood is compared to a baseline lexicon model, which takes only single-word dependencies into account.
0
The computing time is low, since no reordering is carried out.
The evaluation compares the performance of the system with that of several human judges and inter-human agreement on a single correct way to segment a text.
0
(1992).
However, using the top-level semantic classes of WordNet proved to be problematic as the class distinctions are too coarse.
0
Consequently, we cut their evidence values in half to lessen their influence.
Bean and Riloff also used bootstrapping to extend their semantic compatibility model, proposed using caseframe network for anaphora resolution, information extraction patterns to identify contextual clues for determining compatibility between NPs.
0
To resolve the anaphor, we survey the final belief values assigned to each candidate’s singleton set.
The second algorithm builds on a boosting algorithm called AdaBoost.
0
This paper discusses the use of unlabeled examples for the problem of named entity classification.
The bias of automatic methods in favour of statistical systems seems to be less pronounced on out-of-domain test data.
0
Given the closeness of most systems and the wide over-lapping confidence intervals it is hard to make strong statements about the correlation between human judgements and automatic scoring methods such as BLEU.
BABAR's performance in both domains of terrorism and natural disaster, and the contextual-role knowledge in pronouns have shown successful results.
0
In our coreference resolver, we define θ to be the set of all candidate antecedents for an anaphor.
However, these approaches are ill-equipped to directly represent type-based constraints such as sparsity.
0
5.1 Data Sets.
Nevertheless, only a part of this corpus (10 texts), which the authors name "core corpus", is annotated with all this information.
0
When the signal is a coordinating conjunction, the second span is usually the clause following the conjunction; the first span is often the clause preceding it, but sometimes stretches further back.
The overall performance of their joint framework demonstrated that a probability distribution obtained over mere syntactic contexts using a Treebank grammar and a data-driven lexicon outperformed upper bounds proposed by previous joint disambiguation systems and achieved segmentation and parsing results on a par with state-of-the-art standalone applications results.
0
Such discrepancies can be aligned via an intermediate level of PoS tags.
It is annotated with several data: morphology, syntax, rhetorical structure, connectors, correference and informative structure.
0
It reads a file with a list of German connectives, and when a text is opened for annotation, it highlights all the words that show up in this list; these will be all the potential connectives.
The second algorithm builds on a boosting algorithm called AdaBoost.
0
AdaBoost is given access to a weak learning algorithm, which accepts as input the training examples, along with a distribution over the instances.
This paper offers a broad insight into of Arabic constituency parsing by analyzing the interplay of linguistic phenomena, annotation choices, and model design.
0
But Rehbein and van Genabith (2007) showed that Evalb should not be used as an indication of real difference— or similarity—between treebanks.
Foster et all describe a new approach to SMT adaptation that weights out-of-domain phrase pairs according to their relevance to the target domain, determined by both how similar to it they appear to be, and whether they belong to general language or not.
0
The natural baseline (baseline) outperforms the pure IN system only for EMEA/EP fren.
Finally, several coreference systems have successfully incorporated anaphoricity determination modules.
0
If no candidate satisfies this condition (which is often the case), then the anaphor is left unresolved.
The contextual rules are restricted and may not be applicable to every example, but the spelling rules are generally applicable and should have good coverage.
0
We also removed the context-type feature type when using the CoBoost approach.
they evaluate various specific aspects of the segmentation, as well as the overall segmentation performance.
0
An examination of the subjects' bracketings confirmed that these instructions were satisfactory in yielding plausible word-sized units.
Across eight European languages, their approach results in an average absolute improvement of 10.4% over a state-of-the-art baseline, and 16.7% over vanilla hidden Markov models induced with the Expectation Maximization algorithm.
0
For a sentence x and a state sequence z, a first order Markov model defines a distribution: (9) where Val(X) corresponds to the entire vocabulary.
The resulting model is compact, efficiently learnable and linguistically expressive.
0
The difference between the featureless model (+PRIOR) and our full model (+FEATS) is 13.6% and 7.7% average error reduction on best and median settings respectively.
It is well-known that English constituency parsing models do not generalize to other languages and treebanks.
0
Diacritics can also be used to specify grammatical relations such as case and gender.
Bean and Riloff also used bootstrapping to extend their semantic compatibility model, proposed using caseframe network for anaphora resolution, information extraction patterns to identify contextual clues for determining compatibility between NPs.
0
For each caseframe, BABAR collects the semantic classes associated with the head nouns of NPs that were extracted by the caseframe.
This architecture provides a uniform framework in which it is easy to incorporate not only listed dictionary entries but also morphological derivatives, and models for personal names and foreign names in transliteration.
0
The first is an evaluation of the system's ability to mimic humans at the task of segmenting text into word-sized units; the second evaluates the proper-name identification; the third measures the performance on morphological analysis.
This architecture provides a uniform framework in which it is easy to incorporate not only listed dictionary entries but also morphological derivatives, and models for personal names and foreign names in transliteration.
0
It is important to bear in mind, though, that this is not an inherent limitation of the model.
One can trivially create situations in which strictly binary-branching trees are combined to create a tree with only the root node and the terminal nodes, a completely flat structure.
0
Similarly Figures 1 and 2 show how the isolated constituent precision varies by sentence length and the size of the span of the hypothesized constituent.
This paper discusses the Potsdam Commentary Corpus, a corpus of german assembeled by potsdam university.
0
All commentaries have been annotated with rhetorical structure, using RSTTool4 and the definitions of discourse relations provided by Rhetorical Structure Theory (Mann, Thompson 1988).
The evaluation compares the performance of the system with that of several human judges and inter-human agreement on a single correct way to segment a text.
0
Chang of Tsinghua University, Taiwan, R.O.C., for kindly providing us with the name corpora.
This topic has been getting more attention, driven by the needs of various NLP applications.
0
Before explaining our method in detail, we present a brief overview in this subsection.
While many systems had similar performance, the results offered interesting insights, especially, about the relative performance of statistical and rule-based systems.
0
The test data was again drawn from a segment of the Europarl corpus from the fourth quarter of 2000, which is excluded from the training data.
Due to many similarly performing systems, they are not able to draw strong conclusions on the question of correlation of manual and automatic evaluation metrics.
0
The bias of automatic methods in favor of statistical systems seems to be less pronounced on out-of-domain test data.
The evaluation compares the performance of the system with that of several human judges and inter-human agreement on a single correct way to segment a text.
0
Clearly this is not the only way to estimate word-frequencies, however, and one could consider applying other methods: in partic­ ular since the problem is similar to the problem of assigning part-of-speech tags to an untagged corpus given a lexicon and some initial estimate of the a priori probabilities for the tags, one might consider a more sophisticated approach such as that described in Kupiec (1992); one could also use methods that depend on a small hand-tagged seed corpus, as suggested by one reviewer.
It is well-known that English constituency parsing models do not generalize to other languages and treebanks.
0
64 94.
The use of global features has shown excellent result in the performance on MUC-6 and MUC-7 test data.
0
If is not initCaps, then (not-initCaps, ) is set to 1.
They have made use of local and global features to deal with the instances of same token in a document.
0
In (Bikel et al., 1997) and (Bikel et al., 1999), performance was plotted against training data size to show how performance improves with training data size.
Here we present two algorithms.
0
Pseudo-code describing the generalized boosting algorithm of Schapire and Singer is given in Figure 1.
The use of global features has shown excellent result in the performance on MUC-6 and MUC-7 test data.
0
For example, if Barry is not in commonWords and is found in the list of person first names, then the feature PersonFirstName will be set to 1.
They have made use of local and global features to deal with the instances of same token in a document.
0
, Sun day, then the feature DayOfTheWeek is set to 1.
The code is open source, has minimal dependencies, and offers both C++ and Java interfaces for integration.
0
Therefore, performance is more closely tied to the underlying data structure than to the cache.
They proposed a single joint model for performing both morphological segmentation and syntactic disambiguation which bypasses the associated circularity.
0
This essentially means that their model does not result in a true joint inference, as executions for different tasks involve tuning a parameter separately.
The contextual rules are restricted and may not be applicable to every example, but the spelling rules are generally applicable and should have good coverage.
0
The algorithm builds two classifiers in parallel from labeled and unlabeled data.
Here we present two algorithms.
0
This intuition is born out by the experimental results.
The fact that projective dependency parsers can never exactly reproduce the analyses found in non-projective tree banks is often neglected because of the relative scarcity of problematic constructions.
0
Unlike Kahane et al. (1998), we do not regard a projectivized representation as the final target of the parsing process.
The use of global features has shown excellent result in the performance on MUC-6 and MUC-7 test data.
0
Such constraints are derived from training data, expressing some relationship between features and outcome.
The authors cluster NE instance pairs based on the words in the context using bag-of-words methods.
0
Although this is not a precise criterion, most cases we evaluated were relatively clear-cut.
A large number of rules are needed for coverage of the domain, suggesting that a fairly large number of labeled examples should be required to train a classifier.
0
The approach uses both spelling and contextual rules.
They showed the efficacy of graph-based label propagation for projecting part-of-speech information across languages.
0
We define a symmetric similarity function K(uZ7 uj) over two foreign language vertices uZ7 uj E Vf based on the co-occurrence statistics of the nine feature concepts given in Table 1.
they showed that the average agreement among the human judges is .76, and the average agreement between ST(system) and the humans is .75, or about 99% of the interhuman agreement.
0
The major problem for our seg­ menter, as for all segmenters, remains the problem of unknown words (see Fung and Wu [1994]).
Replacing this with a ranked evaluation seems to be more suitable.
0
Figure 1 provides some statistics about this corpus.