text stringlengths 82 736 | label int64 0 1 |
|---|---|
a 5-gram language model was created with the sri language modeling toolkit and trained using the gigaword corpus and english sentences from the parallel data---a 4-gram language model was trained on the target side of the parallel data using the srilm toolkit from stolcke | 1 |
we perform pre-training using the skipgram nn architecture available in the word2vec tool---to get a dictionary of word embeddings , we use the word2vec tool 2 and train it on the chinese gigaword corpus | 1 |
ambiguity is a problem for the vector representation scheme used here , because the two components of an ambiguous vector can add up in a way that makes it by chance similar to an unambiguous word of a different syntactic category---ambiguity is the task of building up multiple alternative linguistic structures for a single input | 1 |
we use the stanford pos tagger to obtain the lemmatized corpora for the parss task---we use the sri language model toolkit to train a 5-gram model with modified kneser-ney smoothing on the target-side training corpus | 0 |
we find that proper-nouns constitute 40 % of query terms , and proper nouns and nouns together constitute over 70 % of query terms---we show that the majority of query terms are proper nouns , and the majority of queries are noun-phrases , which may explain the success of this data | 1 |
the feature weights were tuned on the wmt newstest2008 development set using mert---the features were tuned using mert on the wmt 2012 tuning sets | 1 |
merlo and stevenson presented an automatic classification of three types of english intransitive verbs , based on argument structure and heuristics to thematic relations---merlo and stevenson classify a smaller number of 60 english verbs into three verb classes , by utilising supervised decision trees | 1 |
all features were log-linearly combined and their weights were optimized by performing minimum error rate training---the nnlm weights are optimized as the other feature weights using minimum error rate training | 1 |
we apply our model to the english portion of the conll 2012 shared task data , which is derived from the ontonotes corpus---we use the datasets , experimental setup , and scoring program from the conll 2011 shared task , based on the ontonotes corpus | 1 |
word embeddings are used in many natural language processing tasks---word embeddings have been used to help to achieve better performance in several nlp tasks | 1 |
culotta and sorensen , 2004 ) extended this work to calculate kernels between augmented dependency trees---this tree kernel was slightly generalized by culotta and sorensen to compute similarity between two dependency trees | 1 |
we use different pretrained word embeddings such as glove 1 and fasttext 2 as the initial word embeddings---we use pre-trained 50 dimensional glove vectors 4 for word embeddings initialization | 1 |
relation extraction is the problem of populating a target relation ( representing an entity-level relationship or attribute ) with facts extracted from natural-language text---relation extraction ( re ) is the task of recognizing relationships between entities mentioned in text | 1 |
since the training data is imbalanced , we specifically designed a two-step classifier to address subtask a---due to the imbalanced characteristic of the training data , we specifically adopted a two-step classifier to deal with subtask a | 1 |
in this study , we address the problem of extracting relations between entities from wikipedia¡¯s english articles---study is intended to deal with the problem of extracting binary relations between entity pairs from wikipedia ¡¯ s english version | 1 |
we apply online training , where model parameters are optimized by using adagrad---we train our neural model with stochastic gradient descent and use adagrad to update the parameters | 1 |
we also explore the neural network with few features using n-gram bi-lstms---we also explore bi-lstm models to avoid the detailed feature engineering | 1 |
these features are the output from the srilm toolkit---we implemented scaling , which is similar to that for hmms , in the forward-backward phase of crf training to deal with very long sequences due to sentence concatenation | 0 |
semantic role labeling ( srl ) is the task of automatically labeling predicates and arguments in a sentence with shallow semantic labels---semantic role labeling ( srl ) is a kind of shallow semantic parsing task and its goal is to recognize some related phrases and assign a joint structure ( who did what to whom , when , where , why , how ) to each predicate of a sentence ( cite-p-24-3-4 ) | 1 |
the assumption is that a word vector is learned in such a way that it best predicts its surrounding words in a sentence or a document---we have augmented our algorithm to handle the compilation of weighted rules into weighted finite-state transducers | 0 |
hosseini et al solve single step or multistep homogeneous addition and subtraction problems by learning verb categories from the training data---this paper presents an email importance corpus annotated through amazon | 0 |
the setting itself , specifically , transductive svms , was first introduced by vapnik---the tsvm , a representative of transductive inference method , was introduced by joachims | 1 |
wang et al propose a topical n-gram model that automatically determines unigram words and phrases based on context , and assigns a mixture of topics to both individual words and n-gram phrases---for example , the topical n-gram model introduced by wang et al models unigram and n-gram phrases as mixture of topics based on the nearby word context | 1 |
following previous works on semantic noun classification , we used grs as features for noun clustering---following previous semantic noun classification experiments , we use the grammatical relations as features for clustering | 1 |
in this paper , we address the above challenges with a framework of matrix co-factorization---in this paper , we introduced a framework of matrix co-factorization | 1 |
coreference resolution is the task of partitioning the set of mentions of discourse referents in a text into classes ( or ‘ chains ’ ) corresponding to those referents ( cite-p-12-3-14 )---coreference resolution is the task of grouping mentions to entities | 1 |
svms are a new learning method but have been reported by joachims to be well suited for learning in text classification---svms have been shown to be robust in classification tasks involving text where the dimensionality is high | 1 |
to obtain these features , we use the word2vec implementation available in the gensim toolkit to obtain word vectors with dimension 300 for each word in the responses---to represent the semantics of the nouns , we use the word2vec method which has proven to produce accurate approximations of word meaning in different nlp tasks | 1 |
stroppa et al added souce-side context features to a phrase-based translation system , including conditional probabilities of the same form that we use---stroppa et al add source-side contextual features into a phrase based smt system by integrating context dependent phrasal translation probabilities learned using a decision-tree classifier | 1 |
erkan and radev proposed lexpagerank to compute the sentence saliency based on the concept of eigenvector centrality---erkan and radev proposed a multi-document summarization method using the pagerank algorithm to extract important sentences | 1 |
we built a trigram language model with kneser-ney smoothing using kenlm toolkit---we trained a 3-gram language model on all the correct-side sentences using kenlm | 1 |
in this work , we apply the neural network models to the pun location task---in this paper , we focus on the task of pun location , which aims to identify the pun word | 1 |
text categorization is a classical text information processing task which has been studied adequately ( cite-p-18-1-9 )---for the word-embedding based classifier , we use the glove pre-trained word embeddings | 0 |
we adopt a neural crf with a long-short-termmemory feature layer for baseline pos tagger---we use the long short-term memory architecture for recurrent layers | 1 |
perkins et al , 2003 ) reported that l 1 -norm should be chosen for a problem where most given features are irrelevant---perkins et al , 2003 ) reported that l1-regularizer should be chosen for a problem where most of given features are irrelevant | 1 |
based on these two additional corpora and with l3 as the pivot language , we build a word alignment model for l1 and l2---in live chats , wu et al and forsyth defined 15 dialogue acts for casual online conversations based on previous sets and characteristics of conversations | 0 |
transliteration is the process of converting terms written in one language into their approximate spelling or phonetic equivalents in another language---phonetic translation across these pairs is called transliteration | 1 |
goldwater and mcclosky use morphological analysis on the czech side to get improvements in czech-to-english statistical machine translation---goldwater and mcclosky show improvements in a czech to english word-based translation system when inflectional endings are simplified or removed entirely | 1 |
we used the moses toolkit for performing statistical machine translation---we conducted baseline experiments for phrasebased machine translation using the moses toolkit | 1 |
in this paper we will consider sentence-level approximations of the popular bleu score---in this article we give lower-case bleu scores , except in section 6 where we investigate the effect of different recasing models | 1 |
the skip-gram and continuous bag-of-words models of mikolov et al propose a simple single-layer architecture based on the inner product between two word vectors---we complement the neural approaches with a simple neural network that uses word representations , namely a continuous bag-of-words model | 1 |
for all models , we use l 2 regularization and run 100 epochs of adagrad with early stopping---baldwin and li evaluate the effect of different normalization actions on dependency parsing performance for the social media domain | 0 |
coreference resolution is a complex problem , and successful systems must tackle a variety of non-trivial subproblems that are central to the coreference task — e.g. , mention/markable detection , anaphor identification — and that require substantial implementation efforts---coreference resolution is the problem of identifying which mentions ( i.e. , noun phrases ) refer to which real-world entities | 1 |
for efficiency , we follow the hierarchical softmax optimization used in word2vec---we use the popular word2vec 1 tool proposed by mikolov et al to extract the vector representations of words | 1 |
we estimate the parameters by maximizingp using the expectation maximization algorithm---for this purpose , we turn to the expectation maximization algorithm | 1 |
lda is a generative model that learns a set of latent topics for a document collection---lda is a topic model that generates topics based on word frequency from a set of documents | 1 |
the proposed approach trains models based on only a part of the training set that is more similar to the target domain---moreover , these models are domain-specific , and their performance drops substantially when they are used in a new domain | 0 |
for automatic parsing , we made use of the wellknown charniak parser---for generating con-stituency trees , we used the charniak parser | 1 |
we have shown that a general conversation summarization approach can achieve results on par with state-of-the-art systems that rely on features specific to more focused domains---yu and dredze proposed a model to learn word embeddings based on lexical relations of words from wordnet and ppdb | 0 |
we used the phrasebased translation system in moses 5 as a baseline smt system---we adapted the moses phrase-based decoder to translate word lattices | 1 |
in addition , we use l2 regularization and dropout technique to build a robust system---we used the srilm toolkit to create 5-gram language models with interpolated modified kneser-ney discounting | 0 |
distributional models , on the other hand , use statistics on contextual data from large corpora to predict semantic similarity of words and phrases---distributional models use statistics of word cooccurrences to predict semantic similarity of words and phrases , based on the observation that semantically similar words occur in similar contexts | 1 |
we use the glove word vector representations of dimension 300---we use the glove vector representations to compute cosine similarity between two words | 1 |
the pool is 747 blog sentences 5 from the balanced corpus of contemporary written japanese---these sentence examples were blog articles in the balanced corpus of contemporary written japanese core data | 1 |
this can be solved by the km algorithm for maximum matching in a bipartite graph---this maximum weighted bipartite matching problem can be solved in otime using the kuhnmunkres algorithm | 1 |
we used the svd implementation provided in the scikit-learn toolkit---for nb and svm , we used their implementation available in scikit-learn | 1 |
mutalik et al developed negfinder , a rule-based system that recognises negated patterns in medical documents---mutalik et al developed another rule based system called negfinder that recognizes negation patterns in biomedical text | 1 |
semantic role labeling ( srl ) is a task of automatically identifying semantic relations between predicate and its related arguments in the sentence---semantic role labeling ( srl ) is the task of identifying the predicate-argument structure of a sentence | 1 |
recurrent neural networks are another natural choice to model text due to their capability of processing arbitrary-length sequences---we obtain significant improvements on answer selection and dialogue act analysis | 0 |
semantic role labeling ( srl ) is the process of extracting simple event structures , i.e. , “ who ” did “ what ” to “ whom ” , “ when ” and “ where ”---in a relatively high-dimensional feature space may suffer from the data sparseness problem and thus exhibit less discriminative power on unseen data | 0 |
word sense disambiguation ( wsd ) is a key enabling technology that automatically chooses the intended sense of a word in context---the process of identifying the correct meaning , or sense of a word in context , is known as word sense disambiguation ( wsd ) | 1 |
the log-linear feature weights are tuned with minimum error rate training on bleu---these models can be tuned using minimum error rate training | 1 |
as the grammar is based on a monostratal theory of grammar , annotation by manual disambiguation determines syntactic and semantic structure at the same time---as the grammar is based on a monostratal theory of grammar it is possible to simultaneously annotate syntactic and semantic structure without overburdening the annotator | 1 |
we use pre-trained glove vector for initialization of word embeddings---we also use glove vectors to initialize the word embedding matrix in the caption embedding module | 1 |
for feature building , we use word2vec pre-trained word embeddings---we pre-train the word embedding via word2vec on the whole dataset | 1 |
the output was evaluated against reference translations using bleu score which ranges from 0 to 1---the system output is evaluated using the meteor and bleu scores computed against a single reference sentence | 1 |
the irstlm toolkit is used to build ngram language models with modified kneser-ney smoothing---the target language model is built on the target side of the parallel data with kneser-ney smoothing using the irstlm tool | 1 |
in this paper , we propose to incorporate internal information for lexical sememe prediction---in this paper , we introduced character-level internal information for lexical sememe prediction | 1 |
sentiment classification is the fundamental task of sentiment analysis ( cite-p-15-3-11 ) , where we are to classify the sentiment of a given text---this model was evaluated on the switchboard corpus tors | 0 |
we used srilm to build a 4-gram language model with interpolated kneser-ney discounting---we build an open-vocabulary language model with kneser-ney smoothing using the srilm toolkit | 1 |
first , it includes several features based on large span continuous space language models that have already proved their efficiency both for the translation task and the quality estimation task---the translation quality is evaluated by case-insensitive bleu and ter metrics using multeval | 0 |
the confidence model produces a score based on several predictor features including asr scores , nl scores , and domain knowledge---the target language model is built on the target side of the parallel data with kneser-ney smoothing using the irstlm tool | 0 |
pang et al used a bagof-features framework to train these models from a corpus of movie reviews labelled as positive or negative---we use the sri language model toolkit to train a 5-gram model with modified kneser-ney smoothing on the target-side training corpus | 0 |
once again , segmentation is the part of the process where the automatic algorithms most seriously underperform---the decoder uses a cky-style parsing algorithm to integrate the language model scores | 0 |
keyphrase extraction is a basic text mining procedure that can be used as a ground for other , more sophisticated text analysis methods---keyphrase extraction is the problem of automatically extracting important phrases or concepts ( i.e. , the essence ) of a document | 1 |
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---we trained a trigram language model on the chinese side , with the srilm toolkit , using the modified kneser-ney smoothing option | 1 |
in this paper we address the problem of adapting classifiers trained on the source data and available as black boxes---in this paper we address the domain adaptation scenario without access to source data | 1 |
in this paper , we propose rlie-a3c which uses a3c-based parallel asynchronous agents for training---in this paper , we proposed rlie-a3c , an asynchronous deep reinforcement learning ( rl ) algorithm | 1 |
we used a phrase-based smt model as implemented in the moses toolkit---in the translation tasks , we used the moses phrase-based smt systems | 1 |
we used the svm implementation provided within scikit-learn---in this work , we have proposed a discriminative model for unsupervised morphological | 0 |
kalchbrenner et al proposed to extend cnns max-over-time pooling to k-max pooling for sentence modeling---kalchbrenner et al proposed a dynamic convolution neural network with multiple layers of convolution and k-max pooling to model a sentence | 1 |
however , some studies have shown that adrs are under-estimated due to the fact that they are reported by voluntary reporting systems---however , several studies have shown that adrs are under-estimated because many healthcare professionals do not have enough time to use the adr reporting systems | 1 |
we used a 5-gram language model with modified kneser-ney smoothing implemented using the srilm toolkit---we use srilm toolkit to train a trigram language model with modified kneser-ney smoothing on the target side of training corpus | 1 |
to reduce the potential semantic deviation , we allow for a straightforward perturbation method by replacing nouns and verbs by their synonyms in wordnet---to overcome this problem , we use wordnet to find semantically equivalent replacements for unknown words | 1 |
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---to calculate language model features , we train traditional n-gram language models with ngram lengths of four and five using the srilm toolkit | 1 |
the most prominent of such resources is the framenet , which provides a set of more than 1,200 generic semantic frames , as well as over 200,000 annotated sentences in english---they discriminate learners ’ proficiency adequately trained on error patterns extracted from an esl corpus , and can generate exclusive distractors with taking context of a given sentence into consideration | 0 |
relation extraction is the task of finding relationships between two entities from text---relation extraction is the task of recognizing and extracting relations between entities or concepts in texts | 1 |
this can be seen as a paraphrase identification problem between student answers and reference answers---in this paper , we describe a method for assessing student answers , modeled as a paraphrase identification problem | 1 |
lodhi et al presented a string kernel which measures the similarity between two sentences , or two documents in general , as the number of character subsequences shared between them---lodhi et al described a convolution string kernel , which measures the similarity between two strings by recursively computing matching of all possible subsequences of the two strings | 1 |
in this paper , we propose an interactive group creating system for twitter---morfessor is a commonly used system for unsupervised morphological segmentation | 0 |
sentence compression is the task of generating a grammatical and shorter summary for a long sentence while preserving its most important information---sentence compression is a complex paraphrasing task with information loss involving substitution , deletion , insertion , and reordering operations | 1 |
we use word2vec as the vector representation of the words in tweets---we present a corpus of texts with readability judgments from adults with id ; ( 2 ) we propose a set of cognitively-motivated features which operate at the discourse level | 0 |
all of our parsing models are based on the transition-based dependency parsing paradigm---for all experiments , we used a 5-gram english language model trained on the afp and xinua portions of the gigaword v3 corpus with modified kneser-ney smoothing | 0 |
to sidestep this problem , we employ a variant of importance sampling to help increase the target vocabulary size---to address this problem , we use the approach presented in , which is based on importance sampling | 1 |
for a bigram language model , the decipherment problem is equivalent to the quadratic assignment problem and is np-hard---we show that decipherment using a unigram language model corresponds to solving a linear sum assignment problem | 1 |
srilm toolkit was used to create up to 5-gram language models using the mentioned resources---a 3-gram language model was trained from the target side of the training data for chinese and arabic , using the srilm toolkit | 1 |
for word embeddings , we trained a skip-gram model over wikipedia , using word2vec---to obtain the vector representation of words , we used the google word2vec 1 , an open source tool | 1 |
in the second category , the context of subjective text is used---for more details about the meaning of these labels , see cleuren et al | 0 |
we follow previous works in using gold standard segmentation---we follow previous studies , conducting experiments by using the rst discourse treebank | 1 |
markov models were trained with modified kneser-ney smoothing as implemented in srilm---the two language models were done using the srilm employing linear interpolation and modified k-n discounting | 1 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.