text stringlengths 82 736 | label int64 0 1 |
|---|---|
however , conventional evaluation metrics do not significantly penalize such word order mistakes---from this point of view , conventional automatic evaluation metrics of translation quality disregard word order mistakes | 1 |
the selectivity filter here turned out to be much more time efficient , though---here turned out to be much more time efficient , though | 1 |
part-of-speech ( pos ) tagging is a well studied problem in these fields---part-of-speech ( pos ) tagging is a fundamental natural-language-processing problem , and pos tags are used as input to many important applications | 1 |
cui et al proposed a system utilizing fuzzy relation matching guided by statistical models---evaluated on a news headline dataset , our model yielded higher accuracy | 0 |
optimization with regard to the bleu score is done using minimum error rate training as described in venugopal et al---optimization with regard to the bleu score is done using minimum error rate training as described by venugopal et al | 1 |
this paper deals with an application of automatic titling---this paper presents an original application consisting in titling | 1 |
we build domain-dependent semantic confidence models to improve the rejection of unreliable slu results---domain-dependent confidence models significantly improves performance | 1 |
we used the moses mt toolkit with default settings and features for both phrase-based and hierarchical systems---we preprocessed the corpus with tokenization and true-casing tools from the moses toolkit | 1 |
sentiment classification is the task of identifying the sentiment polarity of a given text , which is traditionally categorized as either positive or negative---high quality word embeddings have been proven helpful in many nlp tasks | 0 |
djuric et al use a paragraph2vec approach to classify language on user comments as abusive or clean---we use glove vectors with 200 dimensions as pre-trained word embeddings , which are tuned during training | 0 |
regarding svm we used linear kernels implemented in svm-light---we use the linear kernel 6 svm , as our text classifier | 1 |
we train a 4-gram language model on the xinhua portion of english gigaword corpus by srilm toolkit---we further used a 5-gram language model trained using the srilm toolkit with modified kneser-ney smoothing | 1 |
reordering is a result of a given derivation , and cyk-based decoding used in tree-based approaches is more syntax-aware than the simple pbsmt decoding algorithm---coreference resolution is the task of partitioning a set of mentions ( i.e . person , organization and location ) into entities | 0 |
we use the same annotation scheme as to model decision-making dialogue---we use the same annotation scheme as in order to model decision-making dialogue | 1 |
semantic role labeling ( srl ) is the task of automatically annotating the predicate-argument structure in a sentence with semantic roles---we use the mallet implementation of conditional random fields | 0 |
wordnet is a general english thesaurus which additionally covers biological terms---wordnet is a manually created lexical database that organizes a large number of english words into sets of synonyms ( i.e . synsets ) and records conceptual relations ( e.g. , hypernym , part of ) among them | 1 |
the pre-processed monolingual sentences will be used by srilm or berkeleylm to train a n-gram language model---a 4-gram language model was trained on the target side of the parallel data using the srilm toolkit from stolcke | 1 |
based on the motivations , we propose a tagging scheme accompanied with the endto-end model to settle this problem---in this paper , we propose a novel tagging scheme and investigate the endto-end models | 1 |
chambers and jurafsky present a system which learns narrative chains from newswire texts---chambers and jurafsky proposed a narrative chain model based on scripts | 1 |
the most challenging work of cross-lingual language modeling is to solve the syntactic discrepancies between the source and target languages---cross-lingual language modeling can be easily applied to speech translation of other language pairs | 1 |
if the name of a target entity has a disambiguation page in wikipedia , we have two or more candidate reference pages---this paper presents the results of a study of the correlation between named entities ( people , places , or organizations | 0 |
in this paper , we name the problem of choosing the correct word from the homophone set the homophone problem---in this paper , we name the problem of choosing the correct word from the homophone set | 1 |
first , we train smt systems with two phrase tables using multiple decoding paths , and combine them in a loglinear model , following koehn and schroeder---first , we interpolate language models from in-domain and out-of-domain data , following koehn and schroeder | 1 |
we train the parameters of the stages separately using adagrad with the perceptron loss function---we optimize the objective by initializing the parameters 胃 to zero and running adagrad | 1 |
out-of-vocabulary ( oov ) words or phrases still remain a challenge in statistical machine translation---words or phrases still remain a challenge in statistical machine translation | 1 |
this similarity score is used to find the nearest neighbors of the test sentence from the training data---in particular , collobert et al and turian et al learn word embeddings to improve the performance of in-domain pos tagging , named entity recognition , chunking and semantic role labelling | 0 |
translation performance was measured by case-insensitive bleu---in this paper we propose to address the problem of automatic labelling of latent topics learned from twitter | 0 |
berland and charniak proposed a system for part-of relation extraction , based on the approach---berland and charniak used similar pattern-based techniques and other heuristics to extract meronymy relations | 1 |
we also use a 4-gram language model trained using srilm with kneser-ney smoothing---we further used a 5-gram language model trained using the srilm toolkit with modified kneser-ney smoothing | 1 |
visweswariah et al and tromble and eisner have considered the source reordering problem to be a problem of learning word reordering from word-aligned data---we tokenize and frequent-case the data with the standard scripts from the moses toolkit | 0 |
pang and lee use a graph-based technique to identify and analyze only subjective parts of texts---pang and lee apply text-categorization techniques to the subjective portions of the sentiment document | 1 |
recommendations expressed in this paper are those of the authors and do not necessarily the views of the sponsors---in this paper are those of the authors and do not necessarily the views of the sponsors | 1 |
in this work , we chose to start with criteria related to content choice---in the project , we focus on content-related criteria | 1 |
feng and cohn propose a word-based markov model to integrate translation and reordering into one model and use the sophisticated hierarchical pitman-yor process which backs off from larger to smaller context to provide dynamic adaptive smoothing---feng and cohn also utilize a markov model for mt , but learn the parameters through a more sophisticated estimation technique that makes use of pitman-yor hierarchical priors | 1 |
we introduce the tree-lstm , a generalization of lstms to tree-structured network topologies---in this paper , we introduce a generalization of the standard lstm architecture to tree-structured network topologies | 1 |
we use a minibatch stochastic gradient descent algorithm together with an adagrad optimizer---in this paper , three subclasses of lfg ' s called nc-lfg ' s , dc-lfg ' s and fc-lfg ' s are introduced and the generative capacities of the above mentioned | 0 |
for evaluation , caseinsensitive nist bleu is used to measure translation performance---we use a list of such connectives compiled by and study the statistics of our corpus to discover the discourse relations | 0 |
consequently , pos induction is a vibrant research area ( see section 2 )---pos induction is a popular topic and several studies ( cite-p-13-1-4 ) have been performed | 1 |
we use a weighted synchronous context free grammar , which was previously used in chiang for hierarchical phrase-based machine translation---relation extraction is a fundamental step in many natural language processing applications such as learning ontologies from texts ( cite-p-12-1-0 ) and question answering ( cite-p-12-3-6 ) | 0 |
for the automatic evaluation we used the bleu and meteor algorithms---we evaluated the system using bleu score on the test set | 1 |
we use 100-dimension glove vectors which are pre-trained on a large twitter corpus and fine-tuned during training---to train our neural algorithm , we apply word embeddings of a look-up from 100-d glove pre-trained on wikipedia and gigaword | 1 |
this paper discusses the differences and relations of six dimensions of subjectivity---sections discuss characteristics and relations of the six dimensions of subjectivity | 1 |
sentiment analysis is the natural language processing ( nlp ) task dealing with the detection and classification of sentiments in texts---sentiment analysis is a fundamental problem aiming to give a machine the ability to understand the emotions and opinions expressed in a written text | 1 |
by grouping opinion holders of different stances on diverse social and political issues , we can gain better understanding of the relationships among countries or among organizations---by grouping opinion holders of different stance on diverse social and political issues , we can a have better understanding of the relationships among countries or among organizations | 1 |
besides , chinese is a topic-prominent language , the subject is usually covert and the usage of words is relatively flexible---mcdonald and pereira presented a graph-based parser that can generate graphs in which a word may depend on multiple heads , and evaluated it on the danish treebank | 0 |
the international corpus of learner en-glish was widely used until recently , despite its shortcomings 2 being widely noted---the above-mentioned international corpus of learner english was widely used until recently , despite its shortcomings 3 being widely noted | 1 |
retrieval effectiveness is found to be strongly influenced by the translation quality of the queries---retrieval effectiveness was found to be strongly influenced by term list size | 1 |
our trigram word language model was trained on the target side of the training corpus using the srilm toolkit with modified kneser-ney smoothing---a 4-gram language model was trained on the target side of the parallel data using the srilm toolkit from stolcke | 1 |
knowledge bases such as freebase and yago play a pivotal role in many nlp related applications---knowledge bases like freebase , dbpedia , and nell are extremely useful resources for many nlp tasks | 1 |
we showed that our keyphrase-based system performs better than a baseline of extracting the sentence with the highest sentiment score---evaluation shows that our sentence extraction method performs better than a baseline of taking the sentence with the strongest sentiment | 1 |
we use the glove vectors of 300 dimension to represent the input words---princeton wordnet is an english lexical database that groups nouns , verbs , adjectives and adverbs into sets of cognitive synonyms , which are named as synsets | 0 |
unification is a central operation in recent computational linguistic research---unification is a basic operation which allows ( a ) to verify if constraints on concatenation are respected ; ( b ) to produce a flow of information between functor and argument | 1 |
twitter is a widely used social networking service---twitter is a very popular micro blogging site | 1 |
we also compare the results with sentiment specific word embeddings , where we use fully connected layers along with attention as the downstream model---word alignments can be applied to acquire synonyms automatically | 0 |
in the application section , we start by presenting an open-source software package for gp modelling in python : gpy---in the application section , we start by presenting an open-source software package for gp modelling | 1 |
lexical simplification is a subtask of text simplification ( cite-p-16-3-3 ) concerned with replacing words or short phrases by simpler variants in a context aware fashion ( generally synonyms ) , which can be understood by a wider range of readers---for the document embedding , we use a doc2vec implementation that downsamples higher-frequency words for the composition | 0 |
wang et al exploit an in-domain language model to score sentences---datasets demonstrate that our joint model significantly outperforms the previous pipelined counterparts , and also achieves better or comparable performance than other approaches to amr parsing , without utilizing external semantic resources | 0 |
we use a random forest classifier , as implemented in scikit-learn---we use the scikit-learn machine learning library to implement the entire pipeline | 1 |
the target language model is trained by the sri language modeling toolkit on the news monolingual corpus---the language model is trained with the sri lm toolkit , on all the available french data without the ted data | 1 |
bengio et al presented a neural network language model where word embeddings are simultaneously learned along with a language model---following , we use the shift-reduce style algorithm to efficiently encode the word aligned phrase-pair as a normalized decomposition tree | 0 |
kilicoglu and bergler showed that manually identified syntactic patterns are effective in classifying sentences as speculative or not---sennrich et al also created synthetic parallel data by translating target-language monolingual text into the source language | 0 |
through our experiments on japanese why-qa , we show that a combination of the above methods can improve why-qa accuracy---by applying these ideas to japanese why-qa , we improved precision by 4 . 4 % against all the questions in our test set | 1 |
related to accommodate this result , we sought to de- to this , it should provide explicit support for velop an architecture that is more general than representing alternative specifications at any a simple pipeline , and thus supports the range point---we created 5-gram language models for every domain using srilm with improved kneserney smoothing on the target side of the training parallel corpora | 0 |
the significance of this work is thus to show that , by working with a simple ¡°knowledge graph¡± representation , we can make a viable version of ¡°interpretation as scene construction¡±---in this paper , we propose a syllable-based method for tweet normalization | 0 |
the language model was constructed using the srilm toolkit with interpolated kneser-ney discounting---lin et al develop a sentence-level recurrent neural network language model that takes a sentence as input and tries to predict the next one based on the sentence history vector | 0 |
mitchell and lapata investigate several vector composition operations for representing short sentences---from the second experiment , we can conclude that taking definition structure into account helps to get better classification | 0 |
unsupervised parsing attracts researchers for many years ,---unsupervised parsing has attracted researchers for decades for recent reviews ) | 1 |
incometo select the most fluent path , we train a 5-gram language model with the srilm toolkit on the english gigaword corpus---end-to-end learning with neural networks has proven to be effective in parsing natural language | 0 |
in this paper we propose a new graph-based method that uses the knowledge in a lkb ( based on wordnet ) in order to perform unsupervised word sense disambuation---in this paper we present a novel graph-based wsd algorithm which uses the full graph of wordnet efficiently , performing significantly better that previously published approaches in english | 1 |
goldwater and griffiths employ a bayesian approach to pos tagging and use sparse dirichlet priors to minimize model size---goldwater and griffiths propose a bayesian approach for learning the hmm structure | 1 |
in recent work , however , we succeed in distinguishing arguments from adjuncts using evidence extracted from a parsed corpus---in recent preliminary work , however , we have succeeded in distinguishing arguments from adjuncts using corpus evidence | 1 |
we use 300-dimensional vectors that were trained and provided by word2vec tool using a part of the google news dataset 4---we use the publicly available 300-dimensional word vectors of mikolov et al , trained on part of the google news dataset | 1 |
this is based on the technique resnik uses for disambiguating noun groups---this is based on a technique for disambiguating noun groups using wordnet by resnik | 1 |
feature weights were set with minimum error rate training on a development set using bleu as the objective function---uchimoto et al also showed that it was possible to replace question-based evaluation with matching of grammatical patterns with no performance loss | 0 |
to train the model , we adopt the averaged perceptron algorithm with early update , following huang and sagae---huang et al described and evaluated a bi-gram hmm tagger that utilizes latent annotations | 0 |
recent work has shown that the effect of eye gaze in facilitating spoken language processing varies among different users---recent works have shown that eye gaze can facilitate spoken language processing in conversational systems | 1 |
in this paper , we deal with the problem of product aspect rating prediction---in this paper , we address the problem of product aspect rating prediction | 1 |
we measured performance using the bleu score , which estimates the accuracy of translation output with respect to a reference translation---for the evaluation of translation quality , we used the bleu metric , which measures the n-gram overlap between the translated output and one or more reference translations | 1 |
our translation model is implemented as an n-gram model of operations using the srilm toolkit with kneser-ney smoothing---we use srilm to build 5-gram language models with modified kneser-ney smoothing | 1 |
we make use of the english dependency treebank , developed on the computational paninian grammar model , for this work---information extraction ( ie ) is the task of extracting information from natural language texts to fill a database record following a structure called a template | 0 |
we train the positive vs negative classifier with liblinear---the sri language modeling toolkit was employed to train a five-gram japanese lm on the training set | 0 |
1 ‘ speakers ’ and ‘ listeners ’ are interchangeably used with ‘ authors ’ and ‘ readers ’ in this artic---1 ‘ speakers ’ and ‘ listeners ’ are interchangeably used with ‘ authors ’ and ‘ readers ’ | 1 |
for math-w-2-6-2-13 , we write math-w-2-6-2-21 to denote the interval math-w-2-6-2-30 , and use [ i ] as a shorthand for math-w-2-6-2-51---later , it have been applied in natural language processing tasks and outperformed traditional models such as bag of words , n-grams and their tfidf variants | 0 |
the most commonly used word embeddings were word2vec and glove---for word embeddings , we used popular pre-trained word vectors from glove | 1 |
sasano , kawahara , and kurohashi conducted similar work with japanese indirect anaphora---sasano et al conducted similar work with japanese indirect anaphora | 1 |
a 4-gram language model was trained on the target side of the parallel data using the srilm toolkit---the pre-processed monolingual sentences will be used by srilm or berkeleylm to train a n-gram language model | 1 |
we compute the interannotator agreement in terms of the bleu score---we evaluate the performance of different translation models using both bleu and ter metrics | 1 |
our trigram word language model was trained on the target side of the training corpus using the srilm toolkit with modified kneser-ney smoothing---we used the srilm toolkit to train a 4-gram language model on the xinhua portion of the gigaword corpus , which contains 238m english words | 1 |
to compare the relative quality of different metrics , we apply bootstrapping re-sampling on the data , and then use paired t-test to determine the statistical significance of the correlation differences---our baseline system is phrase-based moses with feature weights trained using mert | 0 |
the srilm toolkit was used for training the language models using kneser-ney smoothing---in addition , a 5-gram lm with kneser-ney smoothing and interpolation was built using the srilm toolkit | 1 |
a description of the ibm models for statistical machine translation can be found in---the popular ibm models for statistical machine translation are described in | 1 |
feature hashing is a technique of converting string features to vectors---feature hashing is a method for mapping a highdimensional input to a low-dimensional space using hashing | 1 |
our smt-based query expansion techniques are based on a recent implementation of the phrasebased smt framework---our decoder is based on the phrase-based smt model described by koehn et al and implemented , for example , in the popular moses decoder | 1 |
neural networks , working on top of conventional n-gram back-off language models , have been introduced in as a potential means to improve discrete language models---neural networks , working on top of conventional n-gram back-off language models , have been introduced in as a potential means to improve conventional language models | 1 |
we also perform comparison experiments with the partially joint models---at each iteration , the antecedent distribution is used as an attention mechanism to optionally update existing span representations | 0 |
a 4-gram language model is trained on the xinhua portion of the gigaword corpus with the srilm toolkit---the target-side language models were estimated using the srilm toolkit | 1 |
we have used the srilm with kneser-ney smoothing for training a language model for the first stage of decoding---finally , we used kenlm to create a trigram language model with kneser-ney smoothing on that data | 1 |
part-of-speech ( pos ) tagging is the task of assigning each of the words in a given piece of text a contextually suitable grammatical category---turian and melamed observed that uniform example biases bproduced lower accuracy as training progressed , because the induced classifiers minimized the example-wise error | 0 |
newman et al showed that an automated evaluation metric based on word co-occurrence statistics gathered from wikipedia could predict human evaluations of topic quality---the method using pmi proposed by newman et al relies on co-occurrences of words in an external reference source such as wikipedia for automatic evaluation of topic quality | 1 |
lexical functional grammar is a member of the family of constraint-based grammars---lexical functional grammar is a constraint-based , lexicalist approach to the architecture of the grammar | 1 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.