text stringlengths 82 736 | label int64 0 1 |
|---|---|
transliteration is often defined as phonetic translation ( cite-p-21-3-2 )---transliteration is a process of rewriting a word from a source language to a target language in a different writing system using the word ’ s phonological equivalent | 1 |
text categorization is the classificationof documents with respect to a set of predefined categories---text categorization is the task of automatically assigning predefined categories to documents written in natural languages | 1 |
deep convolutional neural networks s are recently extensively used in many computer vision and nlp tasks---we built a 5-gram language model from it with the sri language modeling toolkit | 0 |
furthermore , we train a 5-gram language model using the sri language toolkit---the morphological disambiguation component of our parser is based on more and tsarfaty , modified to accommodate ud pos tags and morphological features | 0 |
in this section , we formulate the sequential decoding problem in the context of perceptron algorithm and crfs---in this section , we propose a parameter estimation procedure for the crfs incorporating partial or ambiguous annotations | 1 |
for the second problem , the model needs to be fed with information on delivery time---in the decoding phase , our model can also generate a numerical value | 1 |
cohn and lapata present a supervised tree-to-tree transduction method for sentence compression---in our experiments , all word vectors are initialized by glove 1 | 0 |
the translation results are evaluated by caseinsensitive bleu-4 metric---the translations are evaluated in terms of bleu score | 1 |
we employ the pretrained word vector , glove , to obtain the fixed word embedding of each word---to keep consistent , we initialize the embedding weight with pre-trained word embeddings | 1 |
various optimisations were made to each string comparison method to reduce retrieval time , of the type described by baldwin and tanaka---evaluation of retrieval accuracy is carried out according to a modified version of the method proposed by baldwin and tanaka | 1 |
videos were then created of all of the schedules for all of the sentences , using the festival speech synthesiser and the ruth animated talking head---we then made videos of every schedule for every sentence , using the festival speech synthesiser and the ruth talking head | 1 |
shen et al proposed a target dependency language model for smt to employ target-side structured information---the language models are 4-grams with modified kneser-ney smoothing which have been trained with the srilm toolkit | 0 |
collobert et al , kalchbrenner et al , and kim use convolutional networks to deal with varying length sequences---kalchbrenner et al introduced a convolutional neural network for sentence modeling that uses dynamic k-max pooling to better model inputs of varying sizes | 1 |
despite the widespread need , the search engines often fail in returning relevant and trustworthy health information---however , it is always difficult for search engines to return relevant and trustworthy health information every time if the symptoms are not accurately described | 1 |
entity linking ( el ) is the task of automatically linking mentions of entities ( e.g . persons , locations , organizations ) in a text to their corresponding entry in a given knowledge base ( kb ) , such as wikipedia or freebase---the grefenstette relation extractor produces context relations that are then lemmatised using the minnen et al morphological analyser | 0 |
we perform our translation experiments using an in-house state-of-the-art phrase-based smt system similar to moses---we use the state-of-the-art phrase-based machine translation system moses perform our machine translation experiments | 1 |
we find that smlkr enables us to recover more systematicity from a lexicon of monomorphemic english words than reported in previous global analyses---on preliminary versions of the monomorphemic lexicon , we noticed that the model detected high degrees of systematicity in words with suffixes | 1 |
in this paper , we explored a new direction to employ the latent meanings of morphological compositions rather than the internal compositions themselves to train word embeddings---incorporating the morphological compositions ( surface forms ) of words , we decide to employ the latent meanings of the compositions ( underlying forms ) to train the word embeddings | 1 |
in this paper , we also follow the same approach for word sense disambiguation---in this line of research , we verify the effectiveness of our approach | 0 |
abstract meaning representation is a semantic representation where the meaning of a sentence is encoded as a rooted , directed graph---abstract meaning representation is a semantic formalism which represents sentence meaning in a form of a rooted directed acyclic graph | 1 |
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 the task of automatic recognition of individual predicates together with their major roles ( e.g . frame elements ) as they are grammatically realized in input sentences | 1 |
past work in relation extraction has focused on binary relations in single sentences---on relation extraction , past work focused primarily on binary relations in single sentences | 1 |
in this paper , we also follow the same approach for word sense disambiguation---using unsupervised methods , this method can be seen as a semi-supervised word sense disambiguation approach | 1 |
on real-world tasks , our method achieves 7 times speedup on citation matching , and 13 times speedup on large-scale author disambiguation---the model weights were trained using the minimum error rate training algorithm | 0 |
we train word embeddings using the continuous bag-of-words and skip-gram models described in mikolov et al as implemented in the open-source toolkit word2vec---to train the link embeddings , we use the speedy , skip-gram neural language model of mikolov et al via their toolkit word2vec | 1 |
the second step is to apply a series of transformations to the parse tree , effectively reordering the surface string on the source language side of the translation system---second step is to apply a series of transformations to the resulting parse tree , effectively reordering the surface string on the source language side of the translation system | 1 |
we evaluate the translation quality using the case-sensitive bleu-4 metric---we also report the results using bleu and ter metrics | 1 |
in this work , we are interested in finding a robust document representation strategy to address the intra-topic variability problem---work is that we proposed a effective variability normalization approach for robust document representation | 1 |
the model was built using the srilm toolkit with backoff and good-turing smoothing---we used a 5-gram language model with modified kneser-ney smoothing implemented using the srilm toolkit | 1 |
we measure translation quality via the bleu score---for this task , we use the widely-used bleu metric | 1 |
a simile is a figure of speech comparing two fundamentally different things---a simile is a figure of speech comparing two essentially unlike things , typically using “ like ” or “ as ” ( cite-p-18-3-1 ) | 1 |
we apply the 3-phase learning procedure proposed by where we first create word embeddings based on the skip-gram model---to convert into a distributed representation here , a neural network for word embedding learns via the skip-gram model | 1 |
we used srilm to build a 4-gram language model with interpolated kneser-ney discounting---we trained a 4-gram language model with kneser-ney smoothing and unigram caching using the sri-lm toolkit | 1 |
the charniak-lease phrase structure parses are transformed into the collapsed stanford dependency scheme using the stanford tools---tees applies the charniak and johnson parser with the mcclosky biomedical model , converting the phrasestructure parses into dependencies using the stanford tools | 1 |
we also used pre-trained word embeddings , including glove and 300d fasttext vectors---for representing words , we used 100 dimensional pre-trained glove embeddings | 1 |
we use adagrad with a batch size of 20 as the optimisation method that automatically adapts the learning rate in training---we used stochastic gradient decent with batch size 100 and adagrad to adapt the learning rate in training | 1 |
to reduce overfitting , we apply the dropout method to regularize our model---we used the opennmt-tf framework 4 to train a bidirectional encoder-decoder model with attention | 0 |
we initialize these word embeddings with glove vectors---like pavlopoulos et al , we initialize the word embeddings to glove vectors | 1 |
coreference resolution is a fundamental component of natural language processing ( nlp ) and has been widely applied in other nlp tasks ( cite-p-15-3-9 )---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 |
lee and seneff , 2008 ) proposed an approach based on pattern matching on trees combined with word n-gram counts for correcting agreement misuse and some types of verb form errors---lee and seneff proposed an approach based on pattern matching on trees combined with word n-gram counts for correcting agreement misuse and some types of verb form errors | 1 |
we build an open-vocabulary language model with kneser-ney smoothing 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 |
we used the sri language modeling toolkit with kneser-kney smoothing---we used srilm to build a 4-gram language model with kneser-ney discounting | 1 |
we use a bidirectional long short-term memory rnn to encode a sentence---for example , vanderwende associated verbs extracted from definitions in an online dictionary with abstract relations | 0 |
word sense disambiguation ( wsd ) is a natural language processing ( nlp ) task in which the correct meaning ( sense ) of a word in a given context is to be determined---lin et al utilize selective attention to aggregate the information of all sentences to extract relational facts | 0 |
li et al adopt a co-training approach which deploys classifiers trained on personal and impersonal view data sets---following , li proposed a co-training approach which exploits subjective and objective views for semi-supervised sentiment classification | 1 |
we use a fourgram language model with modified kneser-ney smoothing as implemented in the srilm toolkit---we trained a trigram language model on the chinese side , with the srilm toolkit , using the modified kneser-ney smoothing option | 1 |
for example , tokens like ¡®iphone¡¯ , ¡®pes¡¯ ( a game name ) and ¡®xbox¡¯ will be considered as nsw , however , these words do not need normalization---an interesting implementation to get the word embeddings is the word2vec model which is used here | 0 |
socher et al introduce a matrix-vector recursive neural network model that learns compositional vector representations for phrases and sentences---later , they proposed a matrix-vector recursive neural network model to learn compositional vector representations for phrases and sentences of any length | 1 |
the translation results are evaluated by caseinsensitive bleu-4 metric---the translation quality is evaluated by case-insensitive bleu-4 | 1 |
the component features are weighted to minimize a translation error criterion on a development set---the log-linear parameter weights are tuned with mert on a development set to produce the baseline system | 1 |
the significance of this work is thus to show that a simple ¡°knowledge graph¡± representation allows a version of ¡°interpretation as scene construction¡± to be made viable---by working with a simple ¡° knowledge graph ¡± representation , we can make a viable version of ¡° interpretation as scene construction ¡± | 1 |
the bleu metric has been widely accepted as an effective means to automatically evaluate the quality of machine translation outputs---a similar idea called ibm bleu score has proved successful in automatic machine translation evaluation | 1 |
key components are entity embeddings , a neural attention mechanism over local context windows , and a differentiable joint inference stage for disambiguation---that combines entity embeddings , a contextual attention mechanism , an adaptive local score combination , as well as unrolled differentiable message passing for global inference | 1 |
the smt systems used a kenlm 5-gram language model , trained on the mono-lingual data from wmt 2015---a 5-gram language model of the target language was trained using kenlm | 1 |
after standard preprocessing of the data , we train a 3-gram language model using kenlm---an english 5-gram language model is trained using kenlm on the gigaword corpus | 1 |
we use the maximum entropy model as a classifier---we used a regularized maximum entropy model | 1 |
luo et al perform the clustering step within a bell tree representation---luo et al propose an approach based on the bell tree to address this problem | 1 |
the gain is a significant reduction in the size number of transformational rules , as much as a factor of three for certain verb classes---from that of prefixes and suffixes , we gain a significant reduction in the number of rules required , as much as a factor of three for certain verb types | 1 |
that is , people tend to act with least effort so as to minimize the cost of energy at both individual and collective levels to the language usage---that is , people tend to act under the least effort in order to minimize the cost of energy at both individual level and collective level to language usage | 1 |
semantic generality is an important indicator of semantic change---meaning change is an important sub-process of innovative meaning change | 1 |
it has been empirically shown that word embeddings could capture semantic and syntactic similarities between words---continuous representations of words have been found to capture syntactic and semantic regularities in language | 1 |
in ( b ) , the templatic grammar improved over the baseline by finding the correct prefix but falsely posited a suffix---in ( b ) , the templatic grammar improved over the baseline by finding the correct prefix | 1 |
dependency parsing is a way of structurally analyzing a sentence from the viewpoint of modification---dependency parsing is a basic technology for processing japanese and has been the subject of much research | 1 |
these lexical chains have many practical applications in ir and computational linguistics such as hypertext construction---we propose a neural architecture which learns a distributional semantic representation that leverage both document and sentence level information | 0 |
we selected conditional random fields as the baseline model---we trained linear-chain conditional random fields as the baseline | 1 |
input layer word embeddings are initialized with glove embeddings pre-trained on twitter text---word embeddings are initialized with glove 27b trained on tweets and are trainable parameters | 1 |
a 4-gram language model was trained on the monolingual data by the srilm toolkit---the target-side language models were estimated using the srilm toolkit | 1 |
our approach relies on several semantic similarity features based on fine-tuned word embeddings and topics similarities---we focus on features that use semantic knowledge such as word embeddings , various features extracted from word embeddings , and topic | 1 |
in this paper , we show how to overcome both limitations---in this work , we address these limitations by enriching the model | 1 |
in this paper , we trade off exact computation for enabling the use of more complex loss functions for coreference resolution ( cr )---in this paper , we study the use of more expressive loss functions in the structured prediction framework for cr , although | 1 |
dependency parsing is a topic that has engendered increasing interest in recent years---dependency parsing is a very important nlp task and has wide usage in different tasks such as question answering , semantic parsing , information extraction and machine translation | 1 |
another popular way to learn word representations is based on the neural language model---one of the most useful neural network techniques for nlp is the word embedding , which learns vector representations of words | 1 |
the charniak-lease phrase structure parses are transformed into the collapsed stanford dependency scheme using the stanford tools---mention properties were obtained from parse trees using the the stanford typed dependency extractor | 1 |
we used moses for pbsmt and hpbsmt systems in our experiments---in this section , we compare 2-layer fcrf with mixed-label lcrf and cross-product lcrf on the joint prediction task | 0 |
arabic is a morphologically complex language---we used a phrase-based smt model as implemented in the moses toolkit | 0 |
in the approach of zettlemoyer and collins , the training data consists of sentences paired with their meanings in lambda form---for representing proper chunks , we employ iob2 representation , one of those which have been studied well in various chunking tasks of natural language processing | 0 |
nominalization is a very productive process---nominalization is a highly productive phenomena | 1 |
conversely , a comparable corpus is a collection of multilingual documents written over the same set of classes ( ni et al. , 2011 ; yogatama and tanaka-ishii , 2009 ) without any restriction about translation or perfect correspondence between documents---a comparable corpus is a collection of texts composed independently in the respective languages and combined on the basis of similarity of content ( cite-p-12-1-15 ) | 1 |
word sense disambiguation ( wsd ) is the process of determining which sense of a homograph is used in a given context---word sense disambiguation ( wsd ) is the process of assigning a meaning to a word based on the context in which it occurs | 1 |
these language models were built up to an order of 5 with kneser-ney smoothing using the srilm toolkit---a 4-gram language model was trained on the target side of the parallel data using the srilm toolkit | 1 |
therefore , for both chinese and english srl systems , we use the 3-best parse trees of berkeley parser and 1-best parse trees of bikel parser and stanford parser as inputs---we provide the monolingual srl system with 3-best parse trees of berkeley parser , 1-best parse tree of bikel parser and stanford parser | 1 |
for example , for the oov mention “ lukebryanonline ” , our model can find similar mentions like “ thelukebryan ” and “ lukebryan ”---mention “ lukebryanonline ” , our model can find similar mentions like “ thelukebryan ” and “ lukebryan ” | 1 |
topic modelling is a popular statistical method for clustering documents---neural networks , working on top of conventional n-gram back-off language models , have been introduced in as a potential means to improve discrete language models | 0 |
ppi information is critical in understanding biological processes---information is very critical in understanding biological processes | 1 |
twitter is a widely used microblogging environment which serves as a medium to share opinions on various events and products---twitter is a communication platform which combines sms , instant messages and social networks | 1 |
the cosine similarity is based on a distributional model constructed with the word2vec tool and the french corpus frwac---we use the long short-term memory architecture for recurrent layers | 0 |
we trained a 4-gram language model on this data with kneser-ney discounting using srilm---for language model , we used sri language modeling toolkit to train a 4-gram model with modified kneser-ney smoothing | 1 |
the smt systems are tuned on the dev development set with minimum error rate training using bleu accuracy measure as the optimization criterion---recently , convolutional neural networks are reported to perform well on a range of nlp tasks | 0 |
accordingly , we use an adaptive recurrence mechanism to learn a dynamic node representation through attention structure---we use different pretrained word embeddings such as glove 1 and fasttext 2 as the initial word embeddings | 0 |
to reflect this observation , in this paper we explore the value-based formulation approach for arbitrary slot filling tasks---in this paper , we proposed an arbitrary slot filling method that directly deals with the posterior probability of slot values | 1 |
we used google pre-trained word embedding with 300 dimensions---we used the pre-trained google embedding to initialize the word embedding matrix | 1 |
somasundaran and wiebe developed a baseline for stance classification using features based on modal verbs and sentiments---for example , somasundaran and wiebe developed a baseline for stance detection by modeling verbs and sentiments | 1 |
ahmed et al , 2011 ) model news storyline clustering by applying a topic model to the clusters , while simultaneously generating single-linkage clusters using the recurrent chinese restaurant process---we used the scikit-learn library the svm model | 0 |
extensive experiments have validated the effectiveness of the corpus-based method for classifying the word¡¯s sentiment polarity---automatic evaluation metrics , such as the bleu score , were crucial ingredients for the advances of machine translation technology in the last decade | 0 |
moreover , our method employs predicate inversion and repetition to resolve the problem that japanese has a predicate at the end of a sentence---in this work , we detailed the multiple choice questions in subject history of gaokao , present two different approaches to address them | 0 |
the system used a tri-gram language model built from sri toolkit with modified kneser-ney interpolation smoothing technique---the language models are 4-grams with modified kneser-ney smoothing which have been trained with the srilm toolkit | 1 |
recent wsi methods were evaluated under the framework of semeval-2007 wsi task---the collocational wsi approach was evaluated under the framework and corpus of semeval-2007 wsi task | 1 |
kann et al improved the results by on canonical segmentation by applying the encoder-decoder rnn framework---semantic role labeling ( srl ) is the task of automatic recognition of individual predicates together with their major roles ( e.g . frame elements ) as they are grammatically realized in input sentences | 0 |
blacoe and lapata compare different arithmetic functions across multiple representations on a range of compositionality benchmarks---blacoe and lapata compare count and predict representations as input to composition functions | 1 |
relation extraction ( re ) has been defined as the task of identifying a given set of semantic binary relations in text---relation extraction is the task of finding relations between entities in text , which is useful for several tasks such as information extraction , summarization , and question answering ( cite-p-14-3-7 ) | 1 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.