text
stringlengths
82
736
label
int64
0
1
for our purpose we use word2vec embeddings trained on a google news dataset and find the pairwise cosine distances for all words---sentiment classification is a useful technique for analyzing subjective information in a large number of texts , and many studies have been conducted ( cite-p-15-3-1 )
0
language identification is a well-studied problem ( cite-p-20-3-2 ) , but it is typically only studied in its canonical text-classification formulation , identifying a document ’ s language given sample texts from a few different languages---we propose a transition-based parser for spinal parsing , based on the arc-eager strategy
0
related work shows that character ngrams can be successfully applied to detect abusive language in english-language content---we trained a 5-gram sri language model using the corpus supplied for this purpose by the shared task organizers
0
we use a minibatch stochastic gradient descent algorithm together with the adam optimizer---for the loss function , we used the mean square error and adam optimizer
1
from a broad perspective , our approach can be seen as using paraphrases of noun compounds---from a word segmentation perspective , our task can be seen as a case study
1
for the generative model , we used the dependency model with valence as it appears in klein and manning---we use the deterministic harmonic initializer from klein and manning
1
dinu and lapata propose a probabilistic framework for representing word meaning and measuring similarity of words in context---dinu and lapata introduced a probabilistic model for computing word representations in context
1
we induce a topic-based vector representation of sentences by applying the latent dirichlet allocation method---to get the the sub-fields of the community , we use latent dirichlet allocation to find topics and label them by hand
1
turney et al overcome this hurdle by applying a semi-supervised method to quantify noun concreteness---turney et al propose a method that extends a large set of concreteness ratings similar to those in the usf dataset
1
consistent with , we found that more filled pauses in an interviewee responsesegment was a significant indicator of deception---in contrast to , we found that the total duration of an interviewee responsesegment was longer for deceptive speech than for truthful speech
1
we used moses with the default configuration for phrase-based translation---for the phrase based system , we use moses with its default settings
1
during decoding , the nmt decoder enquires the phrase memory and properly generates phrase translations---as ¡® constrained ¡¯ , which used only the provided training and development data
0
to get the the sub-fields of the community , we use latent dirichlet allocation to find topics and label them by hand---to measure the importance of the generated questions , we use lda to identify the important sub-topics from the given body of texts
1
in this research , we use the pre-trained google news dataset 2 by word2vec algorithms---with our structured neural network parser , an improvement of 0 . 6 % over the structured perceptron
0
existing studies have already shown some evidence of the transferability of neural features---studies have shown the effectiveness of neural network-based transfer learning
1
moreover , an indirect comparison indicates that our approach also outperforms previous work based on treebank conversion---extensive experiments show that our approach can effectively utilize the syntactic knowledge from another treebank
1
active learning is a promising way for sentiment classification to reduce the annotation cost---active learning is a framework that makes it possible to efficiently train statistical models by selecting informative examples from a pool of unlabeled data
1
sentiment analysis ( sa ) is a field of knowledge which deals with the analysis of people ’ s opinions , sentiments , evaluations , appraisals , attitudes and emotions towards particular entities ( liu , 2012 )---sentiment analysis ( sa ) is a fundamental problem aiming to allow machines to automatically extract subjectivity information from text ( cite-p-16-5-8 ) , whether at the sentence or the document level ( cite-p-16-3-3 )
1
this approach was pioneered by sch眉tze using second order co-occurrences to construct the context representation---the context clustering approach was pioneered by sch眉tze who used second order co-occurrences to construct the context embedding
1
the results from a crowdsourced survey indicated that news values influence people¡¯s decisions to click on a headline---we further used a 5-gram language model trained using the srilm toolkit with modified kneser-ney smoothing
0
it is suggested that the subtree alignment benefits both phrase and syntax based systems by relaxing the constraint of the word alignment---further experiment shows that the obtained subtree alignment benefits both phrase and syntax based
1
given insufficient training examples , we can improve the pos tagging performance by cross-lingual pos tagging , which exploits affluent pos tagging corpora from other source languages---in the universal dependencies corpus , we show that the proposed transfer learning model improves the pos tagging performance of the target languages
1
in treebanks , empty categories have been used to indicate long-distance dependencies , discontinuous constituents , and certain dropped elements---in practical treebanking , empty categories have been used to indicate long-distance dependencies , discontinuous constituents , and certain dropped elements
1
relation extraction ( re ) is the task of extracting semantic relationships between entities in text---relation extraction is the problem of populating a target relation ( representing an entity-level relationship or attribute ) with facts extracted from natural-language text
1
sometimes , even lexicalized patterns are necessary sun et al extend n-grams to noncontinuous sequential patterns allowing arbitrary gaps between words---sun et al extended n-grams to non continuous sequential patterns allowing arbitrary gaps between words
1
relation extraction is the task of predicting semantic relations over entities expressed in structured or semi-structured text---relation extraction ( re ) is a task of identifying typed relations between known entity mentions in a sentence
1
dependency parsing is the task of building dependency links between words in a sentence , which has recently gained a wide interest in the natural language processing community---dependency parsing is the task to assign dependency structures to a given sentence math-w-4-1-0-14
1
boleda et al presented an approach to regular polysemy where meta-alternations capture regularities in meaning shifts---boleda et al show how distributional models can be used to predict regular meaning alternations for novel words
1
collobert et al , 2011 ) trains a neural network to judge the validity of a given context---collobert et al initially introduced neural networks into the srl task
1
we use the moses toolkit to train our phrase-based smt models---we develop translation models using the phrase-based moses smt system
1
we also used word2vec to generate dense word vectors for all word types in our learning corpus---for a fair comparison to our model , we used word2vec , that pretrain word embeddings at a token level
1
we trained an english 5-gram language model using kenlm---we trained a 3-gram language model on all the correct-side sentences using kenlm
1
coreference resolution is the task of determining when two textual mentions name the same individual---coreference resolution is the task of partitioning a set of entity mentions in a text , where each partition corresponds to some entity in an underlying discourse model
1
relation extraction is the task of finding relationships between two entities from text---relation extraction ( re ) is the task of determining semantic relations between entities mentioned in text
1
semantic role labeling ( srl ) is the task of identifying the predicate-argument structure of a sentence---semantic role labeling ( srl ) is defined as the task to recognize arguments for a given predicate and assign semantic role labels to them
1
furthermore , experiment results show that combining similarity functions from different resources could further improve the performance---and find that , with appropriate term weighting strategy , we are able to exploit the information from lexical resources to significantly improve the retrieval performance
1
incremental parsing is the task of assigning a syntactic structure to an input sentence as it unfolds word by word---incremental parsing is a salient feature of glp
1
we use word2vec 1 toolkit to pre-train the character embeddings on the chinese wikipedia corpus---we use word2vec tool which efficiently captures the semantic properties of words in the corpus
1
hara et al derived turn level ratings from overall ratings of the dialogue which were applied by the users afterwards on a five point scale---i essentially include in the discourse all the components covered by the model of grosz and sidner
0
these methods acquire contextual information directly from unannotated raw text , and senses can be induced from text using some similarity measure---these methods acquire knowledge from unannotated raw text , and induce senses using similarity measures
1
metonymy is a figure of speech , in which one expression is used to refer to the standard referent of a related one ( cite-p-18-1-13 )---in this paper , we present a training method for building a dependency parser for a resource-poor language
0
in this work , we focus on recommendation in internet forums and blogs with discussion threads---in this article , we present a framework to recommend relevant information in internet forums and blogs
1
marcu and wong proposed a phrase-based alignment model which suffered from a massive parameter space and intractable inference using expectation maximisation---marcu and wong describe a joint-probability phrase-based model for alignment , but the approach is limited due to excessive complexity as viterbi inference becomes np-hard
1
experimental results show that the proposed methods are effective to improve the retrieval performance , and their performances are comparable to other top-performing systems in the trec medical records track---we use glove word embeddings , an unsupervised learning algorithm for obtaining vector representations of words
0
we present a bootstrapping framework to automatically create event phrase , agent , and purpose dictionaries---we present a bootstrapping algorithm that automatically acquires event phrases , agent terms , and purpose ( reason )
1
sentiment analysis is the computational analysis of people ’ s feelings or beliefs expressed in texts such as emotions , opinions , attitudes , appraisals , etc . ( cite-p-11-3-3 )---we have developed an open ie system which uses svm tree kernels applied to dependency parses for both tasks
0
davidov et al treated 50 twitter tags and 15 smileys as sentiment labels and a supervised sentiment classification framework was proposed to classify the tweets---davidov et al used 50 hashtags and 15 emoticons as sentiment labels for classification to allow diverse sentiment types for the tweet
1
we use the maximum entropy model for our classification task---we use the mallet implementation of a maximum entropy classifier to construct our models
1
state of the art statistical parsers are trained on manually annotated treebanks that are highly expensive to create---current state-of-the-art statistical parsers are trained on large annotated corpora such as the penn treebank
1
collobert and weston , in their seminal paper on deep architectures for nlp , propose a multilayer neural network for learning word embeddings---in the context of neural modeling for nlp , the most notable work was proposed by collobert and weston , which aims at solving multiple nlp tasks within one framework by sharing common word embeddings
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---coreference resolution is the next step on the way towards discourse understanding
1
we used the basic travel expression corpus , a collection of conversational travel phrases for korean and english---rte is a binary classification task , whose goal is to determine , whether for a pair of texts t and h the meaning of h is contained in t ( cite-p-9-1-3 )
0
relation extraction is a key step towards question answering systems by which vital structured data is acquired from underlying free text resources---relation extraction ( re ) is the task of recognizing relationships between entities mentioned in text
1
to support reproducibility of our results , we publish the yags test set annotations and our frame identification system for research purposes---for the out-of-domain testing of framenet srl ; we publish the annotations for the yags benchmark set and our frame identification system for research purposes
1
word sense disambiguation ( wsd ) is the task of determining the correct meaning or sense of a word in context---a 4-gram language model was trained on the target side of the parallel data using the srilm toolkit
0
once the words are embedded in a continuous space , we treat each question as a boew---we represent a question as a bag-of-embedded-words ( boew ) in a continuous space
1
we will also compare the effects of using different ontologies for ontology-based representations---and thus we hypothesize ontology-based representation may facilitate obtaining better content
1
in philosophy and linguistics , it is generally accepted that negation conveys positive meaning---in philosophy and linguistics , it is accepted that negation conveys positive meaning
1
we measure translation quality via the bleu score---we used a 5-gram language model with modified kneser-ney smoothing , built with the srilm toolkit
0
we start by briefly reviewing the latent dirichlet allocation model---to test this hypothesis , we use a latent dirichlet allocation model
1
our learning method is also inspired by the structured perceptron and its application to incremental parsing---our overall approach is closely related to the discriminative incremental parsing framework of collins and roark
1
in this work , we do isolate pp attachments from other parsing decisions---in this paper , we show that word vector representations can yield significant pp attachment
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 show the description of umcc _ dlsi- ( ddi ) system , which is able to detect and classify drugs in biomedical texts with acceptable efficacy
0
a simile is a form of figurative language that compares two essentially unlike things ( cite-p-20-3-11 ) , such as “ jane swims like a dolphin ”---this paper proposes a novel framework for a large-scale , accurate acquisition method for monolingual semantic knowledge , especially for semantic
0
in practical terms , we will use a paraphrase ranking task derived from the semeval 2007 lexical substitution task---consider the occurrence of verb shed in the following semeval 2007 lexical substitution task
1
our system uses the domain-specific data as one dataset to build a robust system---our participation is to build a generic system that is robust
1
recently , bahdanau et al presented a neural attention model for machine translation and showed that the attention mechanism is helpful for addressing long sentences---recently , rockt盲schel et al adapted an attentional lstm model to textual entailment , and a similar model has been applied to cqa
1
the parse trees are generated using the stanford parser---the base pcfg uses simplified categories of the stanford pcfg parser
1
in section 3 , we explain how to use these gazetteers as features in an ne tagger---in section 3 , we explain how to use these gazetteers as features
1
we show that lvegs can subsume latent variable grammars and compositional vector grammars as special cases---erkan and radev introduced a stochastic graph-based method , lexrank , for computing the relative importance of textual units for multi-document summarization
0
we use the mallet implementation of conditional random fields---we rely on conditional random fields 1 for predicting one label per reference
1
thus we can use the community emotion as signals to detect community-related events---we propose an event detection algorithm based on the sequence of community level emotion
1
relation extraction ( re ) is the task of determining semantic relations between entities mentioned in text---relation extraction is the task of tagging semantic relations between pairs of entities from free text
1
the birnn is implemented with lstms for better long-term dependencies handling---gru is reported to be better for long-term dependency modeling than the simple rnn
1
for the task of event trigger prediction , we train a multi-class logistic regression classifier using liblinear---for all machine learning results , we train a logistic regression classifier implemented in scikitlearn with l2 regularization and the liblinear solver
1
we used a phrase-based smt model as implemented in the moses toolkit---we use the opensource moses toolkit to build a phrase-based smt system
1
in this paper , we mainly propose to use an attention-based cnn-lstm model for this task---we use srilm train a 5-gram language model on the xinhua portion of the english gigaword corpus 5th edition with modified kneser-ney discounting
0
yang et al , 2013 ) used neural network-based lexicon and alignment models inside the hmm alignment model , but they model alignments using a simple distortion model that has no dependence on lexical context---statistical significance in bleu differences was tested by paired bootstrap re-sampling
0
in this paper , we address semantic parsing in a multilingual context---we use stanford parser to perform text processing
0
by employing the graph iteration algorithm proposed in , we can compute the rank of a vertex in the entire graph---stephens et al propose 17 classes targeted to relations between genes
0
wiebe et al analyze linguistic annotator agreement statistics to find bias , and use a similar model to correct labels---wiebe et al train a sentence-level probabilistic classifier on data from the wsj to identify subjectivity in these sentences
1
we use evaluation metrics similar to those in---we use the same evaluation criterion as described in
1
and while discourse parsing is a document level task , discourse segmentation is done at the sentence level , assuming that sentence boundaries are known---we represent terms using pre-trained glove wikipedia 6b word embeddings
0
our experiments confirmed that our method was effective---in our system was particularly effective
1
cite-p-25-1-5 applied a graph-based semi-supervised learning algorithm by ( cite-p-25-3-19 )---we use the svm implementation from scikit-learn , which in turn is based on libsvm
0
the probabilistic language model is constructed on google web 1t 5-gram corpus by using the srilm toolkit---a 4-gram language model was trained on the target side of the parallel data using the srilm toolkit
1
for parsing , only a set of possible vns has to be provided---that is incompatible with the annotated labels of the original sentences , we retrofit the lm with a label-conditional architecture
0
in this paper , we introduce a method that automatically builds text classifiers in a new language by training on already labeled data in another language---in this paper , we propose a new approach to cltc , which trains a classification model in the source language
1
we evaluated translation quality using uncased bleu and ter---we evaluated the translation quality using the bleu-4 metric
1
we used the moses machine translation decoder , using the default features and decoding settings---coreference resolution is the problem of identifying which mentions ( i.e. , noun phrases ) refer to which real-world entities
0
the robust processing capabilities of the parser are demonstrated in its use in improving the accuracy of a speech recognizer---robust processing capabilities of the parser have also been shown to be able to provide a small but significant increase in the accuracy of a speech recognizer
1
we trained the syntax-based system on 751,088 german-english translations from the europarl corpus---we further used a 5-gram language model trained using the srilm toolkit with modified kneser-ney smoothing
0
further , it consistently performs better than monolingual bootstrapping---bootstrapping also does better than monolingual bootstrapping
1
pang et al used supervised learning methods and achieved promising results with simple unigram and bi-gram features---pang et al conducted early polarity classification of reviews using supervised approaches
1
su et al , 2008 ) used heterogeneous relations to find implicit sentiment associations among words---su et al presented a clustering method that utilizes the mutual reinforcement associations between features and opinion words
1
our unconstrained system used word embeddings as additional resources---we used the google news pretrained word2vec word embeddings for our model
1
the approach learns from a small , annotated corpus and the task includes resolving not just pronouns but general noun phrases---for evaluation metric , we used bleu at the character level
0
to deal with this problem , proposed using another objective function to promote diversity in responses---we define the position set of math-w-7-11-0-40 , denoted by math-w-7-11-0-44 , as the set of all positions math-w-7-11-0-53
0
semantic role labeling ( srl ) is the task of automatically annotating the predicate-argument structure in a sentence with semantic roles---semantic role labeling ( srl ) is a kind of shallow sentence-level semantic analysis and is becoming a hot task in natural language processing
1
where language model reduction is required , we apply stolcke entropy pruning to m 1 under the relative perplexity threshold 胃---we then use entropy-based pruning of the language model under a relative perplexity threshold of 胃 to reduce the size of m 1
1