text stringlengths 82 736 | label int64 0 1 |
|---|---|
the language model component uses the srilm lattice-tool for weight assignment and nbest decoding---the srilm toolkit and the htk toolkit are used for generating the lms and computing the wer respectively | 1 |
in this paper , we propose an approach for mining query subtopics from query log---using manually defined similarity measures , our approach produces more desirable query subtopics | 1 |
we used a 5-gram language model with modified kneser-ney smoothing implemented using the srilm toolkit---we build a 9-gram lm using srilm toolkit with modified kneser-ney smoothing | 1 |
we implemented linear models with the scikit learn package---we implement logistic regression with scikit-learn and use the lbfgs solver | 1 |
we use the stanford parser to derive the trees---for parsing , we use the stanford parser | 1 |
the process of creating amr ’ s for sentences is called amr parsing---amr parsing is a much harder task in that the target vocabulary size is much larger , while the size of the dataset is much smaller | 1 |
we use the skip-gram model , trained to predict context tags for each word---we have also shown that the contributions of syntax and discourse information are cumulative | 0 |
markov logic is a probabilistic extension of finite first-order logic---we use the moses toolkit to train various statistical machine translation systems | 0 |
in this paper , we propose a semi-supervised technique that uses sentiment bearing word embeddings to produce a continuous ranking among adjectives that share common semantics---in this paper , we have given a technique that uses sentiment specific word embeddings ( sswe ) to produce a fine-grained intensity | 1 |
we also used word2vec to generate dense word vectors for all word types in our learning corpus---we used word2vec , a powerful continuous bag-of-words model to train word similarity | 1 |
in this paper , we propose a goal-directed random walk algorithm to resolve the above problems---in this paper , we introduce a goal-directed random walk algorithm to increase efficiency of mining | 1 |
for other languages , we use the conll-x multilingual dependency parsing shared task corpora which include gold pos tags---rush et al , 2010 ) introduced dual decomposition as a framework for deriving inference algorithms for serious combinatorial problems in nlp | 0 |
particularly relevant to our work is prior art on predicting code-switch points and language identification---most closely related to our work is the study by solorio and liu who predict code-switching in recorded english-spanish conversations | 1 |
then we perform automatic pos tagging using stanford pos tagger---we use the stanford nlp pos tagger to generate the tagged text | 1 |
five-gram language model parameters are estimated using kenlm---five-gram language models are trained using kenlm | 1 |
we use the adagrad algorithm to optimize the conditional , marginal log-likelihood of the data---we train our neural model with stochastic gradient descent and use adagrad to update the parameters | 1 |
typically , the lexicon models used in statistical machine translation systems are only single-word based , that is one word in the source language corresponds to only one word in the target language---typically , the lexicon models used in statistical machine translation systems do not include any kind of linguistic or contextual information | 1 |
our approach represents subtree-based features on the original gold-standard data to retrain parsers---the language model is trained and applied with the srilm toolkit | 0 |
semantic parsing is the task of mapping natural language to a formal meaning representation---semantic parsing is the task of mapping natural language sentences to a formal representation of meaning | 1 |
the distance measures structural difference of two sentences relative to an existing model---distance is proposed to measure the difference of two sentences | 1 |
zeng et al use a convolutional deep neural network to extract lexical features learned from word embeddings and then fed into a softmax classifier to predict the relationship between words---zeng et al proposed a cnn network integrating with position embeddings to make up for the shortcomings of cnn missing contextual information | 1 |
turney and littman determined the polarity of sentiment words by estimating the point-wise mutual information between sentiment words and a set of seed words with strong polarity---turney and littman manually selected seven positive and seven negative words as a polarity lexicon and proposed using pointwise mutual information to calculate the polarity of a word | 1 |
lin et al and zhang et al propose neural attention schemes to select those informative instances---lin et al utilize selective attention to aggregate the information of all sentences to extract relational facts | 1 |
we map the pos labels in the conll datasets to the universal pos tagset---accordingly , we map the bw tagset which is the output of the madamira tools to the universal tagset | 1 |
for feature building , we use word2vec pre-trained word embeddings---in our word embedding training , we use the word2vec implementation of skip-gram | 1 |
translation performance was measured by case-insensitive bleu---the translation outputs were evaluated with bleu and meteor | 1 |
we propose a differentiable probabilistic framework for querying a database given the agent ’ s beliefs over its fields ( or slots )---in this work , we propose a probabilistic framework for computing the posterior distribution of the user target | 1 |
then , we use word embedding generated by skip-gram with negative sampling to convert words into word vectors---we use word vectors produced by the cbow approach-continuous bagof-words | 1 |
we also obtain the embeddings of each word from word2vec---luong and manning , 2016 ) proposes a hybrid architecture for nmt that translates mostly at the word level and consults the character components for rare words when necessary | 0 |
different types of architectures such as feedforward neural networks and recurrent neural networks have since been used for language modeling---recurrent neural networks have recently achieved state of the art results in natural language processing tasks such as language modeling , parsing , and machine translation | 1 |
the language model is a 5-gram with interpolation and kneser-ney smoothing---relation extraction is a crucial task in the field of natural language processing ( nlp ) | 0 |
we use 300-dimensional glove vectors trained on 6b common crawl corpus as word embeddings , setting the embeddings of outof-vocabulary words to zero---besides concentrating on isolated components , a few approaches have emerged that tackle conceptto-text generation | 0 |
kendall¡¯s math-w-11-5-2-1 can be easily used to evaluate the output of automatic systems , irrespectively of the domain or application at hand---while natural language text is a rich source to obtain broad knowledge about the world , compiling trivial commonsense knowledge from unstructured text is a nontrivial feat | 0 |
to tackle this issue , we leverage pretrained word embeddings , specifically the 300 dimension glove embeddings trained on 42b tokens of external text corpora---we use the pre-trained glove 50-dimensional word embeddings to represent words found in the glove dataset | 1 |
translation quality is evaluated by case-insensitive bleu-4 metric---although sequence labeling is the simplest subclass , a lot of real-world tasks are modeled as problems of this simplest subclass | 0 |
the correlated topic model induces a correlation structure between topics by using the logistic normal distribution instead of the dirichlet---ctm uses the logistic normal distribution to replace the dirichlet prior , so it can capture the correlated structure of topics | 1 |
the lexical semantic relationships between word pairs are key features for many nlp tasks---lexical-semantic relations such as hypernymy or meronymy have proven useful in many nlp tasks | 1 |
we use the glove vector representations to compute cosine similarity between two words---we adopt glove vectors as the initial setting of word embeddings v | 1 |
word segmentation is a prerequisite for many natural language processing ( nlp ) applications on those languages that have no explicit space between words , such as arabic , chinese and japanese---therefore , word segmentation is a crucial first step for many chinese language processing tasks such as syntactic parsing , information retrieval and machine translation | 1 |
dense real-valued distributed representations of words known as word embeddings have become ubiquitous in nlp , serving as invaluable features in a broad range of nlp tasks , eg ,---dense , low-dimensional , real-valued vector representations of words known as word embeddings have proven very useful for nlp tasks | 1 |
language models were built using the sri language modeling toolkit with modified kneser-ney smoothing---the language models are 4-grams with modified kneser-ney smoothing which have been trained with the srilm toolkit | 0 |
recently , rnn-based models have been successfully used in machine translation and dialogue systems---lstms have become more popular after being successfully applied in statistical machine translation | 1 |
third and finally , the baselines reported for resnik¡¯s test set were higher than those for the all-words task---with these goals in mind , we cast reg as a density estimation problem | 0 |
relation extraction is a traditional information extraction task which aims at detecting and classifying semantic relations between entities in text ( cite-p-10-1-18 )---relation extraction is a well-studied problem ( cite-p-12-1-6 , cite-p-12-3-7 , cite-p-12-1-5 , cite-p-12-1-7 ) | 1 |
the challenge of modeling this connection is to ground language at the level of relations---challenge of this view is to model grounding at the level of relations | 1 |
recently , distributed representations have been widely used in a variety of natural language processing tasks---distributed representations of words have been widely used in many natural language processing tasks | 1 |
the various models developed are evaluated using bleu and nist---the bleu metric has been used to evaluate the performance of the systems | 1 |
mikolov et al presents a neural network-based architecture which learns a word representation by learning to predict its context words---there are several studies about grammatical error correction using phrase-based statistical machine translation | 0 |
in our experiments , we used the srilm toolkit to build 5-gram language model using the ldc arabic gigaword corpus---morphological analysis is the segmentation of words into their component morphemes and the assignment of grammatical morphemes to grammatical categories and lexical morphemes to lexemes | 0 |
for some , the regular language is a superset of the context-free language , and for others it is a subset---by definition , a language accepted by a finite automaton is called a regular language | 1 |
medlock and briscoe used single words as input features in order to classify sentences from biological articles as speculative or non-speculative based on semi-automatically collected training examples---medlock and briscoe also used single words as input features in order to classify sentences from scientific articles in biomedical domain as speculative or non-speculative | 1 |
we initialize our word vectors with 300-dimensional word2vec word embeddings---we used word2vec to convert each word in the world state , query to its vector representation | 1 |
we apply standard tuning with mert on the bleu score---we compute the interannotator agreement in terms of the bleu score | 1 |
we use liblinear 9 to solve the lr and svm classification problems---accurately identifying events in unstructured text is an important goal | 0 |
we use lists of discourse markers compiled from the penn discourse treebank and from to identify such markers in the text---it was trained on the webnlg dataset using the moses toolkit | 0 |
the domain and language portability of the proposed system is demonstrated by its successful application across three different domains and two languages---the high performance in different domains is a promising indicator for domain and language portability | 1 |
ghosh and veale have found deep neural networks to perform better compared to support vector machines for sarcasm detection---as deep learning techniques gain popularity , ghosh and veale propose a neural network semantic model for sarcasm detection | 1 |
the first precursors to the acld were the fixit query-free search system , the remembrance agent for just-in-time retrieval , and the implicit queries system---experimental results show that the proposed methods are effective to improve the retrieval | 0 |
we trained a standard 5-gram language model with modified kneser-ney smoothing using the kenlm toolkit on 4 billion running words---during the training process we built n-gram language models for use in decoding and rescoring using the kenlm language modelling toolkit | 1 |
a simile is a figure of speech comparing two essentially unlike things , typically using “ like ” or “ as ” ( cite-p-18-3-1 )---a simile is a comparison between two essentially unlike things , such as “ jane swims like a dolphin ” | 1 |
a disadvantage of the log-linear models is that they require cluster computing resources for practical training---previous discriminative models for ccg required cluster computing resources to train | 1 |
we use the moses toolkit to train our phrase-based smt models---we use the moses package to train a phrase-based machine translation model | 1 |
to get a dictionary of word embeddings , we use the word2vec tool 2 and train it on the chinese gigaword corpus---for the embeddings trained on stack overflow corpus , we use the word2vec implementation of gensim 8 toolkit | 1 |
as for a generative parser , dubey et al proposed an unlexicalized pcfg parser that modified pcfg probabilities to condition the existence of a coordinate structure---dubey et al proposed an unlexicalized pcfg parser that modified pcfg probabilities to condition the existence of syntactic parallelism | 1 |
nakagawa et al , 2010 , used a svm with secondorder polynomial kernel and additional features---nakagawa et al , 2010 ) introduced an approach based on crfs with hidden variables with very good performance | 1 |
we used srilm for training the 5-gram language model with interpolated modified kneser-ney discounting ,---we used the srilm toolkit and kneser-ney discounting for estimating 5-grams lms | 1 |
for all experiments , we used a 4-gram language model with modified kneser-ney smoothing which was trained with the srilm toolkit---a trigram language model with modified kneser-ney discounting and interpolation was used as produced by the srilm toolkit | 0 |
the annotation scheme is based on an evolution of stanford dependencies and google universal part-of-speech tags---the ud annotation has evolved by reconstruction of the standford dependencies and it uses a slightly extended version of google universal tag set for part of speech | 1 |
then we extend the seq2seq framework to jointly conduct template reranking and template-aware summary generation---while we extend the seq2seq framework to conduct template reranking and template-aware summary generation | 1 |
in the approach of zettlemoyer and collins , the training data consists of sentences paired with their meanings in lambda form---this goes beyond previous work on semantic parsing such as lu et al or zettlemoyer and collins which rely on unambiguous training data where every sentence is paired only with its meaning | 1 |
we use the selectfrommodel 4 feature selection method as implemented in scikit-learn---we use the logistic regression implementation of liblinear wrapped by the scikit-learn library | 1 |
relation extraction ( re ) is the process of generating structured relation knowledge from unstructured natural language texts---relation extraction ( re ) is the task of extracting semantic relationships between entities in text | 1 |
we used the phrasebased translation system in moses 5 as a baseline smt system---we used moses , a state-of-the-art phrase-based smt model , in decoding | 1 |
popular word embedding techniques have proven useful for analyzing language evolution---word embedding approaches like word2vec or glove are powerful tools for the semantic analysis of natural language | 1 |
liu et al proposed a recursive neural network designed to model the subtrees , and cnn to capture the most important features on the shortest dependency path---for our purpose we use word2vec embeddings trained on a google news dataset and find the pairwise cosine distances for all words | 0 |
we use 300-dimensional glove vectors trained on 6b common crawl corpus as word embeddings , setting the embeddings of outof-vocabulary words to zero---the language model is trained and applied with the srilm toolkit | 0 |
however , what remains unclear is which of these graphs are most helpful for a specific document exploration task---for some of them , it is largely unknown which type of graph is most helpful for a specific exploratory task | 1 |
recently , convolutional neural networks are reported to perform well on a range of nlp tasks---neural networks perform well for many small-scale classification tasks | 1 |
to exploit these kind of labeling constraints , we resort to conditional random fields---we solve this sequence tagging problem using the mallet implementation of conditional random fields | 1 |
training is done through stochastic gradient descent over shuffled mini-batches with the adagrad update rule---weights are optimized by the gradient-based adagrad algorithm with a mini-batch | 1 |
language models were built using the sri language modeling toolkit with modified kneser-ney smoothing---language models were built with srilm , modified kneser-ney smoothing , default pruning , and order 5 | 1 |
relation extraction is a subtask of information extraction that finds various predefined semantic relations , such as location , affiliation , rival , etc. , between pairs of entities in text---in future work , we will try to collect and annotate data for microblogs in other languages | 0 |
for the syntactic analogy and text classification tasks , lmms also surpass all the baselines---as concerns identifying difficult terms , some applications search for them in vocabularies or in specific corpora | 0 |
in this paper , we have presented a novel application of alternating structure optimization ( aso ) to the semantic role labeling ( srl ) task on nombank---in this paper , we explore a number of different auxiliary problems , and we are able to significantly improve the accuracy of the nombank srl task | 1 |
semantic role labeling ( srl ) is the task of identifying the predicate-argument structure of a sentence---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 |
we implemented the different aes models using scikit-learn---we use a random forest classifier , as implemented in scikit-learn | 1 |
coreference resolution is the task of clustering a set of mentions in the text such that all mentions in the same cluster refer to the same entity---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 |
our transition-based parser is based on a study by zhu et al , which adopts the shift-reduce parsing of sagae and lavie and zhang and clark---our parser is based on the shift-reduce parsing process from sagae and lavie and wang et al , and therefore it can be classified as a transition-based parser , | 1 |
we therefore provide a comparison of the analysis drawn in cite-p-11-1-7 with a standard bootstrap implementation---in this paper , we provide further analysis of experiments originally provided in cite-p-11-1-7 , in addition to further investigation | 1 |
the basic idea of context-group discrimination is to induce senses from contextual similarity---recently , distributed word representations using the skip-gram model has been shown to give competitive results on analogy detection | 0 |
this is done by integrating semantic parsing into the syntactic parsing process---the translation results are evaluated with case insensitive 4-gram bleu | 0 |
we follow the pairwise approach to ranking that reduces ranking to a binary classification problem---we use linear ranking functions and transform the ranking problem into a two-class classification problem | 1 |
parsing is the process of mapping sentences to their syntactic representations---parsing is the task of reconstructing the syntactic structure from surface text | 1 |
previous works on stance detection have focused on congressional debates , company-internal discussions , and debates in online forums---previous work has focused on congressional debates , company-internal discussions , and debates in online forums | 1 |
we use 300d glove vectors trained on 840b tokens as the word embedding input to the lstm---we used the srilm toolkit and kneser-ney discounting for estimating 5-grams lms | 0 |
huang et al presented a new neural network architecture which incorporated both local and global document context , and offered an impressive result---huang et al introduced global document context and multiple word prototypes which distinguishes and uses both local and global context via a joint training objective | 1 |
the weights of the word embeddings use the 300-dimensional glove embeddings pre-trained on common crawl data---the word representations are publicly-available 300-dimension glove 4 word vectors trained on 42 billion tokens of web data | 1 |
we tune phrase-based smt models using minimum error rate training and the development data for each language pair---the berkeley framenet is an ongoing project for building a large lexical resource for english with expert annotations based on frame semantics | 0 |
the feature weights are tuned with mert to maximize bleu-4---the weights for these features are optimized using mert | 1 |
we used the phrasebased translation system in moses 5 as a baseline smt system---we perform named entity tagging using the stanford four-class named entity tagger | 0 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.