text stringlengths 82 736 | label int64 0 1 |
|---|---|
specifically , we adopt linear-chain conditional random fields as the method for sequence labeling---named entity recognition is a challenging task that has traditionally required large amounts of knowledge in the form of feature engineering and lexicons to achieve high performance | 0 |
named entity disambiguation is the task of linking an entity mention in a text to the correct real-world referent predefined in a knowledge base , and is a crucial subtask in many areas like information retrieval or topic detection and tracking---named entity disambiguation ( ned ) is the task of determining which concrete person , place , event , etc . is referred to by a mention | 1 |
finally , our system uses multilayer perceptron ( mlp ) to predict event spans---step , we use the multilayer perceptron ( mlp ) to predict event spans | 1 |
table 2 shows the blind test results using bleu-4 , meteor and ter---we use glove 300-dimension embedding vectors pre-trained on 840 billion tokens of web data | 0 |
we use the word2vec vectors with 300 dimensions , pre-trained on 100 billion words of google news---for our purpose we use word2vec embeddings trained on a google news dataset and find the pairwise cosine distances for all words | 1 |
our idea is inspired from the use of bottleneck features obtained using neural networks for training hmm-based speech recognition systems---our framework is inspired by use of bottleneck features obtained from neural networks in hidden markov model based speech recognition | 1 |
these systems were official entries to the semeval-2010 cross-lingual lexical substitution task---two systems submitted as official entries to the semeval-2010 cross-lingual lexical substitution task | 1 |
we use the logistic regression implementation of liblinear wrapped by the scikit-learn library---we train a linear support vector machine classifier using the efficient liblinear package | 1 |
coreference resolution is the task of identifying all mentions which refer to the same entity in a document---although coreference resolution is a subproblem of natural language understanding , coreference resolution evaluation metrics have predominately been discussed in terms of abstract entities and hypothetical system errors | 1 |
we use a maximum entropy classifier with a large number of boolean features , some of which are novel---word alignment is the problem of annotating parallel text with translational correspondence | 0 |
dimensional , pre-trained fasttext embeddings were used to train the svm models---the learning representation relies on fasttext pre-trained word embeddings | 1 |
a 3-gram language model is trained on the target side of the training data by the srilm toolkits with modified kneser-ney smoothing---we use the sri language model toolkit to train a 5-gram model with modified kneser-ney smoothing on the target-side training corpus | 1 |
finally , we investigate why rm adaptation helps smt performance---rm model adaptation will improve smt performace | 1 |
mishne and de rijke , 2006 ) constructed models to predict the levels of various moods according to the language used by bloggers at a giv-en time---biadsy et al present a system that identifies dialectal words in speech and their dialect of origin through the acoustic signals | 0 |
in this section , we describe our self-disclosure topic model , based on the widely used latent dirichlet allocation , which incorporates those approaches---in smt , we propose a coverage-based approach to nmt to alleviate the over-translation and under-translation problems | 0 |
we adapted the moses phrase-based decoder to translate word lattices---we use the moses smt toolkit to test the augmented datasets | 1 |
we have described a dependency-based system 1 for semantic role labeling of english in the propbank framework---we attempt to represent the cross-linguistic similarities that exist in the consonant inventories of the world ’ s languages through a bipartite network | 0 |
we also explore the neural network with few features using n-gram bi-lstms---yang and eisenstein introduced a highly accurate unsupervised normalization model | 0 |
in the future work , we will explore the hierarchical learning strategy using other machine learning approaches besides online classifier learning approaches such as the simple perceptron algorithm applied in this paper---in the future work , we will explore the hierarchical learning strategy using other machine learning approaches besides online classifier learning approaches | 1 |
for significance tests , we use the wilcoxon signed ranks test---the language models are 4-grams with modified kneser-ney smoothing which have been trained with the srilm toolkit | 0 |
guinaudeau and strube describe a graphbased version of the entity grid which models the interaction between entities and sentences as a bipartite graph---barzilay and lapata propose an entity grid model which represents the distribution of referents in a discourse for sentence ordering | 1 |
in our analyses , we show empirically that these learned attention weights correlate strongly with traditional headedness definitions---attention mechanism also learns a task-specific preference for head words , which we empirically showed correlate strongly with traditional headword definitions | 1 |
as the word embeddings , we used the 300 dimension vectors pre-trained by glove 6---in our experiments , all word vectors are initialized by glove 1 | 1 |
to minimize the objective , we use stochastic gradient descent with the diagonal variant of adagrad---we use a minibatch stochastic gradient descent algorithm together with an adagrad optimizer | 1 |
in this paper we describe our participation at semeval-2018 task 3---we describe our participation to the semeval-2018 task | 1 |
we use the 200-dimensional global vectors , pre-trained on 2 billion tweets , covering over 27-billion tokens---we use 300-dimensional word embeddings from glove to initialize the model | 1 |
discourse parsing is a natural language processing ( nlp ) task with the potential utility for many other natural language processing tasks ( webber et al. , 2011 )---discourse parsing is a challenging task and is crucial for discourse analysis | 1 |
since coreference resolution is a pervasive discourse phenomenon causing performance impediments in current ie systems , we considered a corpus of aligned english and romanian texts to identify coreferring expressions---we derive our gold standard from the semeval 2007 lexical substitution task dataset | 0 |
we use a weighted synchronous context free grammar , which was previously used in chiang for hierarchical phrase-based machine translation---mover ’ s distance provides a distance measure that may quantify a facet of language difference | 0 |
the use of generative probabilistic grammars for parsing is well understood---much of the current research into probabilistic parsing is founded on probabilistic contextfree grammars | 1 |
in this paper , we leverage the ilp method as a core component in our summarization system---in this paper , we propose a bigram based supervised method for extractive document summarization | 1 |
sentiment classification has seen a great deal of attention---sentiment classification has advanced considerably since the work | 1 |
luong and manning proposed a hybrid scheme that consults character-level information whenever the model encounters an oov word---luong and manning designed a hybrid character-and word-based encoder to try to solve the out-of-vocabulary problem | 1 |
part of speech ( pos ) tagging is a quite well defined nlp problem , which consists of assigning to each word in a text the proper morphosyntactic tag for the given context---part of speech ( pos ) tagging is the process of marking up words and punctuation characters in a text with appropriate pos labels | 1 |
our experiments show that our model has higher lexical as well as sentential diversity than baseline models---experiments show that our model generates more diverse outputs than baseline models , and also generates more consistently acceptable output than sampling | 1 |
this paper introduces an unsupervised vector approach to disambiguate words in biomedical text using contextual information from the umls---this paper introduces an unsupervised vector approach to disambiguate words in biomedical text | 1 |
the semeval semantic textual similarity tasks are a popular evaluation venue for the sts problem---major progress has been made in this task in recent years , due primarily to the semeval semantic textual similarity task | 1 |
we use different pretrained word embeddings such as glove 1 and fasttext 2 as the initial word embeddings---we initialize the embedding layer using embeddings from dedicated word embedding techniques word2vec and glove | 1 |
soricut and marcu employed a standard bottom-up chart parsing algorithm with syntactic and lexical features to conduct sentencelevel parsing---soricut and marcu use a standard bottomup chart parsing algorithm to determine the discourse structure of sentences | 1 |
the algorithm is formulated using the framework of parsing as deduction , extended with weights---the algorithm is specified by means of deduction rules , following , and can be implemented using standard tabular techniques | 1 |
to deal with this problem , proposed using another objective function to promote diversity in responses---to address this challenge , introduced a new objective function with mmi to penalize too general responses | 1 |
synonym discovery has been an active topic in a variety of language processing tasks---distributional word similarity has long been an active research area | 1 |
automatic processing of web queries is therefore of utmost importance---automatic processing of web queries is important for high-quality information | 1 |
the standard minimum error rate training algorithm was used for tuning---tuning was performed by minimum error rate training | 1 |
we show that by making trivial adaptations , monolingual parsing models can effectively parse code-mixed data---that we proposed , multilingual and interpolation methods are two competitive methods for parsing code-mixed data | 1 |
again we expand the list using the paraphrase database , resulting in a total of 200 signals---our trigram word language model was trained on the target side of the training corpus using the srilm toolkit with modified kneser-ney smoothing | 0 |
we automatically extract more translation pairs using the europarl parallel corpus and select pairs based on the word frequency in the target language---for a fair comparison among target languages , we extract the intersection of the europarl corpus in our three language pairs so that the source side data is identical for all nmt systems | 1 |
for annotation , we used the brat rapid annotation tool---we used the brat annotation tool for annotating the corpus | 1 |
we used minimum error rate training to tune the feature weights for maximum bleu on the development set---we use mt02 as the development set 4 for minimum error rate training | 1 |
gradability is a property of words that identifies different degrees of the quality the word denotes---gradability is a semantic property that allows a word to describe the intensity of a measure in context , and thus enables comparative constructs | 1 |
these language models were built up to an order of 5 with kneser-ney smoothing using the srilm toolkit---we use srilm toolkit to build a 5-gram language model with modified kneser-ney smoothing | 1 |
for our experiments , we used the latent variablebased berkeley parser---for our investigations , we used the berkeley parser as a source of grammar rule clusters | 1 |
we used 100 dimensional glove embeddings for this purpose---we used crfsuite and the glove word vector | 1 |
furthermore , we plan to integrate the proposed interface within an computer-based interactive platform for speech therapy---in this project , we deal with developing an interactive interface to assist speech therapists with constructing individualized speech | 1 |
event extraction is a task in information extraction where mentions of predefined events are extracted from texts---for all three systems , we used the stanford corenlp package to perform lemmatization and pos tagging of the input sentences | 0 |
sentiment analysis ( sa ) is the task of determining the sentiment of a given piece of text---sentiment analysis ( sa ) is the determination of the polarity of a piece of text ( positive , negative , neutral ) | 1 |
we use a fourgram language model with modified kneser-ney smoothing as implemented in the srilm toolkit---the language models are 4-grams with modified kneser-ney smoothing which have been trained with the srilm toolkit | 1 |
we use a minibatch stochastic gradient descent algorithm together with an adagrad optimizer---to learn grsemi-crfs , we employ adagrad , an adaptive stochastic gradient descent method which has been proved successful in similar tasks | 1 |
the only external linguistic resource required by pem is a parallel text of the target language and another arbitrary language---bpng is a parallel text of the target language and an arbitrary other language , known as the pivot language | 1 |
the rest of this paper , specifically section 2 is a brief related work---coreference resolution is the task of grouping all the mentions of entities 1 in a document into equivalence classes so that all the mentions in a given class refer to the same discourse entity | 0 |
our study demonstrates that the composite kernel is very effective for relation extraction---our study illustrates that the composite kernel can effectively capture both flat and structured features | 1 |
we used srilm to build a 4-gram language model with interpolated kneser-ney discounting---we apply sri language modeling toolkit to train a 4-gram language model with kneser-ney smoothing | 1 |
an unpruned , modified kneser-ney-smoothed 4-gram language model is estimated using the kenlm toolkit---both are estimated with the kenlm toolkit using interpolated kneser-ney smoothing | 1 |
we train a 4-gram language model on the xinhua portion of english gigaword corpus by srilm toolkit---we use srilm toolkit to build a 5-gram language model with modified kneser-ney smoothing | 1 |
language models were built with srilm , modified kneser-ney smoothing , default pruning , and order 5---a 4-gram language model is trained on the monolingual data by srilm toolkit | 0 |
predicting characteristics of twitter users , including political party affiliation has been explored---predicting political affiliation and other characteristics of twitter users has been explored | 1 |
information extraction ( ie ) is the process of identifying events or actions of interest and their participating entities from a text---information extraction ( ie ) is the task of generating structured information , often in the form of subject-predicate-object relation triples , from unstructured information such as natural language text | 1 |
dependency parsing and semantic role labeling are two standard tasks in the nlp community---this grammar consists of a lexicon which pairs words or phrases with regular expression functions | 0 |
nowadays , most conversational systems require extensive human annotation efforts in order to be fit for their task---comparison with additional measures always increases the overall reliability of the evaluation | 0 |
a zero pronoun ( zp ) is a gap in a sentence that is found when a phonetically null form is used to refer to a real-world entity---a zero pronoun ( zp ) is a gap in a sentence which refers to an entity that supplies the necessary information for interpreting the gap | 1 |
minimalist grammars are a formalization of minimalist syntax---mgs are a rigorous formalization of minimalist syntax | 1 |
the weights are trained using a procedure similar to on held-out test data---the model weights were trained using the minimum error rate training algorithm | 1 |
neural networks have also been used to learn representations for use in phrase-structure parsing---recently , recursive neural networks have been proposed for syntactic parsing | 1 |
the rule base utility was evaluated within two lexical expansion applications , yielding better results than other automatically constructed baselines and comparable results to wordnet---we used the pb smt system in moses 12 for je and kj translation tasks | 0 |
we stress that this model is not tied to a particular feature dependency graph---we trained a smt system on 10k french-english sentences from the europarl corpus | 0 |
this maximum weighted bipartite matching problem can be solved in otime using the kuhnmunkres algorithm---this combinatorial optimisation problem can be solved in polynomial time through the hungarian algorithm | 1 |
the standard phrase-based machine translation system focuses on finding the most probable target sentence given the source sentence---the phrase-based approach developed for statistical machine translation is designed to overcome the restrictions of many-to-many mappings in word-based translation models | 1 |
noun-compounds hold an implicit semantic relation between their constituents---for preprocessing the corpus , we use the stanford pos-tagger and parser included in the dkpro framework | 0 |
the nnlm weights are optimized as the other feature weights using minimum error rate training---the parameter weights are optimized with minimum error rate training | 1 |
practically , word-level representation has been extensively explored to improve many downstream natural language processing tasks---distributed representations of words have been widely used in many natural language processing tasks | 1 |
a line of work has been proposed to explore the effect of neural network models for constituent parsing---there has been some work on neural networks for constituent based parsing | 1 |
we have introduced the issue of sentence segmentation of impaired speech , and tested the effectiveness of standard segmentation methods on ppa speech samples---we explore whether we can apply standard approaches to sentence segmentation to impaired speech , and compare our results to the segmentation of broadcast news | 1 |
global features are often described using a continuous feature space , such as color histogram in three different color spaces , or textures using gabor and haar wavelets---for a global representation , features are often described using a continuous feature space , such as a color histogram in three different color spaces , or textures using gabor and haar wavelets | 1 |
we apply the evaluation method used to evaluate vector representation of text sequences by le and mikolov---to represent the tweets , we make use of the doc2vec algorithm described in le and mikolov | 1 |
for example , citation structure or rebuttal links , are used as extra information to model agreements or disagreements in debate posts and to infer their labels---hence we use the expectation maximization algorithm for parameter learning | 0 |
we use 300 dimensional glove embeddings trained on the common crawl 840b tokens dataset , which remain fixed during training---we use glove 300-dimension embedding vectors pre-trained on 840 billion tokens of web data | 1 |
however , handcrafted , well-structured taxonomies such as wordnet , opencyc and freebase , which are publicly available , can be incomplete for new or specialized domains---we also tried early update in the learning algorithm | 0 |
barzilay and lapata recently proposed an entity-based coherence model that aims to learn abstract coherence properties , similar to those stipulated by centering theory---this paper describes a system for navigating large collections of information about cultural heritage | 0 |
furthermore , we train a 5-gram language model using the sri language toolkit---we implement an in-domain language model using the sri language modeling toolkit | 1 |
the log-linear parameter weights are tuned with mert on the development set---the weights for the loglinear model are learned using the mert system | 1 |
hazem and morin , 2012 ) propose a method that filters the entries of the bilingual dictionary on the base of a pos-tagging and a domain relevance measure criteria but no improvements have been demonstrated---hazem and morin recently proposed a method that filters the entries of the bilingual dictionary based upon pos-tagging and domain relevance criteria , but no improvements was demonstrated | 1 |
in order to overcome data sparsity , tfba backs-off and jointly factorizes multiple lower-order tensors derived from an extremely sparse higher-order tensor---directly , tfba performs back-off and jointly factorizes multiple lower-order tensors derived out of the higher-order tensor | 1 |
coreference resolution is the task of determining when two textual mentions name the same individual---coreference resolution is the problem of partitioning a sequence of noun phrases ( or mentions ) , as they occur in a natural language text , into a set of referential entities | 1 |
in this study , we propose a cross-lingual representation learning model bidrl which simultaneously learns both the word and document representations in both languages---in this study , we propose a representation learning approach which simultaneously learns vector representations for the texts | 1 |
we also used word2vec to generate dense word vectors for all word types in our learning corpus---we compared sn models with two different pre-trained word embeddings , using either word2vec or fasttext | 1 |
we employed the machine learning tool of scikit-learn 3 , for training the classifier---we trained the five classifiers using the svm implementation in scikit-learn | 1 |
we propose to implement such models within neural network frameworks with structures , in which the merging parameters can be optimized in a principled way , to minimize a well-defined objective---we propose our models in neural network frameworks with structures , in which the merging parameters can be learned in a principled way to optimize a welldefined objective | 1 |
as the operational semantics of natural language applications improves , even larger improvements are possible---as the operational semantics of natural language applications improve , even larger improvements are possible | 1 |
semantic parsing is the task of mapping natural language to machine interpretable meaning representations---semantic parsing is the task of mapping natural language sentences to a formal representation of meaning | 1 |
in re , our investigations are in the paradigm of distant supervision , which facilitates the creation of large albeit noisy training data---in the relation extraction setting , our research has been in the paradigm of learning under distant supervision | 1 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.