text
stringlengths
82
736
label
int64
0
1
we used a 5-gram language model with modified kneser-ney smoothing , built with the srilm toolkit---for language model , we used sri language modeling toolkit to train a 4-gram model with modified kneser-ney smoothing
1
descriptions are transformed into a vector by adding the corresponding word2vec embeddings---the character embeddings are computed using a method similar to word2vec
1
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---to get the the sub-fields of the community , we use latent dirichlet allocation to find topics and label them by hand
0
language models are built using the sri-lm toolkit---the trigram language model is implemented in the srilm toolkit
1
the translation quality is evaluated by case-insensitive bleu-4---the evaluation metric is the case-insensitive bleu4
1
we used kenlm with srilm to train a 5-gram language model based on all available target language training data---we used the sri language modeling toolkit to train a fivegram model with modified kneser-ney smoothing
1
the translation quality is evaluated by case-insensitive bleu-4 metric---the translation results are evaluated with case insensitive 4-gram bleu
1
our translation model is implemented as an n-gram model of operations using the srilm toolkit with kneser-ney smoothing---we have used the srilm with kneser-ney smoothing for training a language model for the first stage of decoding
1
relation extraction is a fundamental task in information extraction---relation extraction is a challenging task in natural language processing
1
coreference resolution is the process of linking together multiple expressions of a given entity---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
our results show that using our proposed sentence compression model in the summarization system can yield significant performance gain in linguistic quality , without losing much performance on the rouge metric---by incorporating this sentence compression model , our summarization system can yield significant performance gain in linguistic quality without losing much rouge results
1
in this paper , we focus on generic document summarization and keyword extraction for single documents---we propose a novel approach to simultaneously document summarization and keyword extraction for single documents
1
sentiment classification is the task of detecting whether a textual item ( e.g. , a product review , a blog post , an editorial , etc . ) expresses a p ositive or a n egative opinion in general or about a given entity , e.g. , a product , a person , a political party , or a policy---sentiment classification is the task of identifying the sentiment polarity of a given text
1
we use word2vec tool for learning distributed word embeddings---we perform pre-training using the skipgram nn architecture available in the word2vec tool
1
the language model was constructed using the srilm toolkit with interpolated kneser-ney discounting---the model was built using the srilm toolkit with backoff and kneser-ney smoothing
1
hatzivassiloglou and mckeown showed how the pattern x and y could be used to automatically classify adjectives as having positive or negative orientation---hatzivassiloglou and mckeown proposed a supervised algorithm to determine the semantic orientation of adjectives
1
we used srilm to build a 4-gram language model with kneser-ney discounting---for the language model , we used srilm with modified kneser-ney smoothing
1
coreference resolution is the process of determining whether two expressions in natural language refer to the same entity in the world---coreference resolution is a key task in natural language processing ( cite-p-13-1-8 ) aiming to detect the referential expressions ( mentions ) in a text that point to the same entity
1
we use the sri language model toolkit to train a 5-gram model with modified kneser-ney smoothing on the target-side training corpus---we use srilm toolkit to train a trigram language model with modified kneser-ney smoothing on the target side of training corpus
1
in this paper , we have proposed a semi-supervised hierarchical topic models , i.e . sshlda , which aims to solve the drawbacks of hlda and hllda while combine their merits---in this paper , we propose a semi-supervised hierarchical topic model which aims to explore new topics automatically
1
underlying the semantic roles approach is a lexicalist assumption , that is , each verb ’ s lexical entry completely encodes ( more formally , projects ) its syntactic and semantic structures---under a lexicalist approach to semantics , a verb completely encodes its syntactic and semantic structures , along with the relevant syntax-to-semantics
1
a 4-gram language model is trained on the monolingual data by srilm toolkit---a kn-smoothed 5-gram language model is trained on the target side of the parallel data with srilm
1
our approach combines a set of hand-written patterns together with a probabilistic model---in this paper , we develop a probabilistic model that uses a set of patterns and tree matching
1
in our study , we express the relation of word co-occurrence in the form of a graph---in our method is constructed based on word co-occurrence
1
for co-occurrence statistical methods , hu and liu proposed a pioneer research for opinion summarization based on association rules---hu and liu proposed a statistical approach to capture object features using association rules
1
the meta-net project aims to ensure equal access to information by all european citizens---given the model parameters and a sentence pair math-w-2-14-1-11 , compute math-w-2-14-1-18
0
word sense disambiguation ( wsd ) is a particular problem of computational linguistics which consists in determining the correct sense for a given ambiguous word---word sense disambiguation ( wsd ) is a key enabling-technology
1
miller et al adapt a probabilistic context-free parser for information extraction by augmenting syntactic labels with entity and relation labels---miller et al propose an integrated statistical parsing technique that augments parse trees with semantic labels denoting entity and relation types
1
multi-task learning via neural networks have been used to model relationships among the correlated tasks---multi-task learning using a related auxiliary task can lead to stronger generalization and better regularized models
1
blitzer et al used structural correspondence learning to train a classifier on source data with new features induced from target unlabeled data---blitzer et al employ the structural correspondence learning algorithm for sentiment domain adaptation
1
word alignment models were first introduced in statistical machine translation---we train a trigram language model with the srilm toolkit
0
systems that jointly annotate syntactic and semantic dependencies were introduced in the past conll-2008 shared task---in particular , the recent shared tasks of conll 2008 tackled joint parsing of syntactic and semantic dependencies
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 a simpler task than constituent parsing , since dependency trees do not have extra non-terminal nodes and there is no need for a grammar to generate them
1
the continuous bag-of-words approach described by mikolov et al is learned by predicting the word vector based on the context vectors---all models have been estimated using publicly available software , moses , and corpora
0
we primarily compared our model with conditional random fields---we trained linear-chain conditional random fields as the baseline
1
combinatory categorial grammars are a linguistically-motivated model for a wide range of language phenomena---combinatory categorial grammar is a syntactic theory that models a wide range of linguistic phenomena
1
phrase chunking is a natural language processing task that consists in dividing a text into syntactically correlated parts of words---phrase chunking is a natural language processing ( nlp ) task that consists in dividing a text into syntactically correlated parts of words
1
a mention is a reference to an entity such as a word or phrase in a document---each mention is a reference to some entity in the domain of discourse
1
other researchers used a word-character hybrid model , which combines dictionary-lookup and character-based modeling of oov words---kruengkrai et al proposed a hybrid model including character-based and word-based features
1
the tnt tagger and the treetagger are used for tagging and lemmatization---treetagger is used for pos tagging and lemmatization
1
the model weights of all systems have been tuned with standard minimum error rate training on a concatenation of the newstest2011 and newstest2012 sets---the weights 位 m in the log-linear model were trained using minimum error rate training with the news 2009 development set
1
semantic parsing is the task of mapping a natural language ( nl ) sentence into a complete , formal meaning representation ( mr ) which a computer program can execute to perform some task , like answering database queries or controlling a robot---semantic parsing is the task of mapping natural language sentences to a formal representation of meaning
1
word sense disambiguation ( wsd ) is the nlp task that consists in selecting the correct sense of a polysemous word in a given context---word sense disambiguation ( wsd ) is a difficult natural language processing task which requires that for every content word ( noun , adjective , verb or adverb ) the appropriate meaning is automatically selected from the available sense inventory 1
1
the previous probabilistic pos models for agglutinative languages have considered only lexical forms of morphemes , not surface forms of words---models for agglutinative languages have considered only lexical forms of morphemes , not surface forms of words
1
proposed by chiang , the hierarchical phrase-based machine translation model has achieved results comparable , if not superior , to conventional phrase-based approaches---the widely-used hierarchical phrase-based translation framework was introduced by chiang and also relies on a simple heuristic for phrase pair extraction
1
we proposed a model based on arbitrary subtrees of dependency trees---we present an alternative model based on subtrees of dependency trees , so as to extract entities
1
in this paper , we proposed a new method for word sense disambiguation that is based on unlabeled data---in this paper , we present a new approach to word sense disambiguation that is based on selective sampling algorithm
1
we train an english language model on the whole training set using the srilm toolkit and train mt models mainly on a 10k sentence pair subset of the acl training set---one uses confusion networks formed along a skeleton sentence to combine translation systems as described in and
0
we use the opensource moses toolkit to build a phrase-based smt system---we use the moses smt toolkit to test the augmented datasets
1
we begin in section 2 by formally describing the directional word alignment problem---to the model of interest , we first introduce directional word alignment
1
the probabilistic language model is constructed on google web 1t 5-gram corpus by using the srilm toolkit---the english side of the parallel corpus is trained into a language model using srilm
1
named entity recognition ( ner ) is the process by which named entities are identified and classified in an open-domain text---named entity recognition ( ner ) is the task of identifying and typing phrases that contain the names of persons , organizations , locations , and so on
1
in medium-resource setting , the performance was 65.06 % , almost the same as that of the baseline ( 64.7 % )---by 3 . 85 % ) , and in medium-resource scenarios the performance was 65 . 06 % ( almost the same as baseline )
1
twitter is a widely used microblogging environment which serves as a medium to share opinions on various events and products---twitter is a famous social media platform capable of spreading breaking news , thus most of rumour related research uses twitter feed as a basis for research
1
in order to estimate the terms f and f the corpus was automatically parsed by cass , a robust chunk parser designed for the shallow analysis of noisy text---verb-object bigrams for the 30 preselected verbs were obtained from the bnc using cass , a robust chunk parser designed for the shallow analysis of noisy text
1
a number of models have been proposed to learn distributed word or phrase representations in order to predict word occurrences given a local context---based on the distributional hypothesis , various methods for word embeddings have been actively studied
1
generic user affect parameters increase the usefulness of these models---sentence compression is the task of producing a summary at the sentence level
0
we used the moses toolkit for performing statistical machine translation---for phrase-based smt translation , we used the moses decoder and its support training scripts
1
we employ a neural method , specifically the continuous bag-of-words model to learn high-quality vector representations for words---based on the distributional hypothesis , we train a skip-gram model to learn the distributional representations of words in a large corpus
1
inversion transduction grammar , or itg , is a wellstudied synchronous grammar formalism---bracketing transduction grammar is a special case of synchronous context free grammar
1
in this paper , we propose to use a transfer learning approach for sentiment analysis ( semeval2018 task 1 )---in this paper , we apply a transfer learning approach , from a model trained on a similar task
1
besides , he et al built a maximum entropy model which combines rich context information for selecting translation rules during decoding---he et al proposes maximum entropy models which combine rich context information for selecting translation rules during decoding
1
more recently eisenstein et al modeled geographic linguistic variation using twitter data---in this paper , we present an approach to enriching high-order feature representations for graph-based dependency parsing models
0
the language models are 4-grams with modified kneser-ney smoothing which have been trained with the srilm toolkit---table 1 presents the results from the automatic evaluation , in terms of bleu and nist test
0
finally , zhou et al proposed a unified transformation-based learning framework on chinese edt---zhou et al proposed a unified transformation-based learning framework and tested it on chinese edt
1
however , to our knowledge , we give the first detailed analysis on spurious ambiguity of word alignment---we propose a novel framework for speech disfluency detection based on integer linear programming ( ilp )
0
the results show that our method outperforms conventional unsupervised object matching methods---the results showed that our method outperformed conventional object matching methods
1
we used the open source moses decoder package for word alignment , phrase table extraction and decoding for sentence translation---for phrase-based smt translation , we used the moses decoder and its support training scripts
1
we use the l2-regularized logistic regression of liblinear as our term candidate classifier---we trained the l1-regularized logistic regression classifier implemented in liblinear
1
lu et al proposed a deep learning method suited for short texts---lu and li proposed a dnn-based matching model for response selection
1
traditionally , a language model is a probabilistic model which assigns a probability value to a sentence or a sequence of words---the language model defined by the expression is named the conditional language model
1
we use the conditional random fields learning algorithm in order to annotate the words with biesto labels---we train conditional random field as a machine learning algorithm to identify the candidate wordforms that need to be normalized
1
le and mikolov applied paragraph information into the word embedding technique to learn semantic representation---they learned text embeddings using the neural language model from le and mikolov and used them to train a binary classifier
1
for nmt , we applied byte pair encoding to split word into subword segments for both source and target languages---we use byte-pair-encoding to achieve openvocabulary translation with a fixed vocabulary of subword symbols
1
following , we use the bootstrap resampling test to do significance testing---we apply statistical significance tests using the paired bootstrapped resampling method
1
bleu is used as a standard evaluation metric---weights are optimized by mert using bleu as the error criterion
1
mada-arz is an egyptian arabic extension of the morphological analysis and disambiguation of arabic---mada-arz is an egyptian arabic extension of the morphological analysis and disambiguation of arabic tool
1
in this paper we will consider sentence-level approximations of the popular bleu score---bleu is a precision based measure and uses n-gram match counts up to order n to determine the quality of a given translation
0
these models are an instance of conditional random fields and include overlapping features---using all constraints , ignoring noise and overlap , results in surprisingly high accuracy , within 2 % of a fully-supervised approach on three of four tasks
0
we use the stanford part of speech tagger to annotate each word with its pos tag---some of our features are based on the part-of-speech tags assigned by the stanford tagger
1
budanitsky and hirst provide an overview of wordnet-based similarity measures---budanitsky and hirst provided a survey of such wordnet-based measures
1
finally , we extend feature noising for structured prediction to a transductive or semi-supervised setting---in this paper , we make a description of our submitted system to the semeval-2018 shared task
0
we evaluate our approaches on eight language pairs with data sizes ranging from 100k to 8m words , and achieve improvements of up to +4.3 bleu , surpassing phrase-based translation in nearly all settings---the technique is applicable to any language pair and does not require especially difficultto-obtain resources
0
in recent years supervised learning approaches have been widely used in coreference resolution task and achieved considerable success---twitter is a rich resource for information about everyday events – people post their tweets to twitter publicly in real-time as they conduct their activities throughout the day , resulting in a significant amount of mundane information about common events
0
word sense disambiguation ( wsd ) is the task of determining the correct meaning for an ambiguous word from its context---word sense disambiguation ( wsd ) is the task of assigning sense tags to ambiguous lexical items ( lis ) in a text
1
we used yamcha as a text chunker , which is based on support vector machine---we used the chunker yamcha , which is based on support vector machines
1
pantel and lin present a clustering algorithm -coined clustering by committee -that automatically discovers word senses from text---pantel and lin , 2002 , introduce a method known as committee based clustering that discovers word senses
1
we extract the named entities from the web pages using the stanford named entity recognizer---we train a 4-gram language model on the xinhua portion of the english gigaword corpus using the srilm toolkits with modified kneser-ney smoothing
0
in particular , we explore unlabeled data to transfer the predictive power of hybrid models to simple sequence models---at test time , we explore unlabeled data to transfer the predictive power of hybrid models to simple sequence or even local classification models
1
however , statistical approaches to combating fake news has been dramatically limited by the lack of labeled benchmark datasets---however , the lack of manually labeled fake news dataset is still a bottleneck for advancing computational-intensive , broad-coverage models
1
the performance of the system for all subtasks in both languages shows substantial improvements in spearman correlation scores over the baseline models provided by task 1 organizers , ranging from 0.03 to 0.23---we adapt the models of mikolov et al and mikolov et al to infer feature embeddings
0
we employ stem as the atomic translation unit to alleviate data spareness---therefore , we employ stem as the atomic translation unit and use affix information to guide translation
1
we use 5-grams for all language models implemented using the srilm toolkit---we use a fourgram language model with modified kneser-ney smoothing as implemented in the srilm toolkit
1
we used the stanford parser to parse the corpus---we used stanford dependency parser for the purpose
1
collobert et al propose avoiding taskspecific engineering by learning features during model training---collobert et al employ a cnn-crf structure , which obtains competitive results to statistical models
1
we use a fourgram language model with modified kneser-ney smoothing as implemented in the srilm toolkit---for language model , we used sri language modeling toolkit to train a 4-gram model with modified kneser-ney smoothing
1
we also use a 4-gram language model trained using srilm with kneser-ney smoothing---the english side of the parallel corpus is trained into a language model using srilm
1
twitter is a fantastic data resource for many tasks : measuring political ( o ’ connor et al. , 2010 ; tumasjan et al. , 2010 ) , and general sentiment ( cite-p-11-1-3 ) , studying linguistic variation ( cite-p-11-3-2 ) and detecting earthquakes ( cite-p-11-3-18 )---twitter is a huge microblogging service with more than 500 million tweets per day from different locations of the world and in different languages ( cite-p-10-1-6 )
1
we use bleu scores as the performance measure in our evaluation---and without an alignment of graphemes and phonemes , we obtained a word accuracy rate of 75 . 3 % for the 5-dimensional german syllable model
0
we use the stanford segmenter 9 for tokenization , treetagger for lemmatization and partof-speech tagging---we use treetagger with the default parameter file for tokenization , lemmatization and annotation of part-of-speech information in the corpus
1