text stringlengths 82 736 | label int64 0 1 |
|---|---|
in this run , we use a sentence vector derived from word embeddings obtained from word2vec---we first encode each word in the input sentence to an m-dimensional vector using word2vec | 1 |
though expressive and accurate , these models fail to fully exploit gpu parallelism , limiting their computational efficiency---while these models are expressive and accurate , they fail to fully exploit the parallelism opportunities of a gpu | 1 |
in this paper , we present initial experiments in the recognition of deceptive language---in this paper , we explored automatic techniques for the recognition of deceptive language | 1 |
it combines a traditional bag-of-words ( bow ) representation with a distributed vector representation created by a cnn , to retrieve semantically equivalent questions---bow-cnn , combines a bag-of-words ( bow ) representation with a distributed vector representation created by a convolutional neural network ( cnn ) | 1 |
following , we distinguish between minimal and composed rules---following galley et al , we distinguish between minimal and composed rules | 1 |
in this section , we evaluate the log-linear model and compare it with the mle based model presented by bannard and callison-burch 6---in order to assess the performance of our model , we compare it to two variants of the models proposed by bannard and callison-burch | 1 |
there has been a considerable amount of work on arabic morphological analysis---for example , jeon et al . ( cite-p-17-1-9 , cite-p-17-1-10 ) compared four different retrieval methods , i . e . vector space model , okapi , language model ( lm ) , and translation-based model | 0 |
the parameter weights are optimized with minimum error rate training---the log linear weights for the baseline systems are optimized using mert provided in the moses toolkit | 1 |
word sense disambiguation ( wsd ) is a task to identify the intended sense of a word based on its context---the process of identifying the correct meaning , or sense of a word in context , is known as word sense disambiguation ( wsd ) | 1 |
we train our svm classifiers using the liblinear package---we use the svm implementation available in the li-blinear package | 1 |
among the measures that can be used for controlled translation , we focus on translation literalness in this paper---we focused on literalness from among the various measures for controlled translation and defined a translation | 1 |
all the feature weights were trained using our implementation of minimum error rate training---we used a trigram language model trained on gigaword , and minimum error-rate training to tune the feature weights | 1 |
we use a cws-oriented model modified from the skip-gram model to derive word embeddings---we use skip-gram with negative sampling for obtaining the word embeddings | 1 |
we train lms with srilm using jelinek-mercer linear interpolation as a smoothing method---for the language model , we used srilm with modified kneser-ney smoothing | 1 |
in the gsm-based communication system , a vad scheme is used to lengthen the battery power through discontinuous transmission when speech-pause is detected---in the gsm-based wireless system , for instance , a vad module is used for discontinuous transmission to save battery power | 1 |
jiang et al pointed out that long utterances are prone to cause asr errors---jiang et al found that users tend to repeat previous utterances in case of asr errors | 1 |
for the word-embedding based classifier , we use the glove pre-trained word embeddings---for this score we use glove word embeddings and simple addition for composing multiword concept and relation names | 1 |
li et al proposed to transfer common lexical knowledge across domains via matrix factorization techniques---li et al proposed to transfer sentiment knowledge from source domain to target domain using nonnegative matrix factorization | 1 |
we used the srilm toolkit to create 5-gram language models with interpolated modified kneser-ney discounting---for this language model , we built a trigram language model with kneser-ney smoothing using srilm from the same automatically segmented corpus | 1 |
bahdanau et al and luong et al proposed the use of attention mechanisms to translate better by considering the context in which particular target words occur with respect to the source contexts---the mcr also integrates the latest version of the wordnet domains , new versions of the base concepts and the top concept ontology , and the sumo ontology | 0 |
from the introspection aspect , luo et al propose to select supportive law articles and use the articles to enhance the charge prediction accuracy---this phenomenon of the reference is called zero anaphora | 0 |
coreference resolution is the process of linking together multiple expressions of a given entity---coreference resolution is a set partitioning problem in which each resulting partition refers to an entity | 1 |
we trained a 4-gram language model on the xinhua portion of gigaword corpus using the sri language modeling toolkit with modified kneser-ney smoothing---a 4-gram language model is trained on the xinhua portion of the gigaword corpus with the srilm toolkit | 1 |
mihalcea et al developed several corpus-based and knowledge-based word similarity measures and applied them to a paraphrase recognition task---mihalcea et al defines a measure of text semantic similarity and evaluates it in an unsupervised paraphrase detector on this data set | 1 |
nallapati et al propose a recurrent neural network-based sequence-to-sequence model for sequential labelling of each sentence in the document---nallapati et al presented a neural sequential model for the extractive summarization of documents | 1 |
we use mstparser 4 for conventional firstorder model and secondorder model---we use the mstparser implementation described in mcdonald et al for feature extraction | 1 |
experiments on five languages showed that the approach can yield significant improvement in tagging accuracy given sufficiently fine-grained label sets---relation extraction is the task of finding semantic relations between entities from text | 0 |
for the semantic language model , we used the srilm package and trained a tri-gram language model with the default goodturing smoothing---we further used a 5-gram language model trained using the srilm toolkit with modified kneser-ney smoothing | 1 |
user : so i have to remove a file to create a file---user : i want to prevent tom from reading my file | 1 |
when involving in neural networks , both zhang and lapata and wang et al employ recurrent neural network and planning to perform generation---zhang and lapata , 2014 ) employed the recurrent neural network as their basis and further considered the global context using convolutional neural network | 1 |
we use the skip-gram model with negative sampling to learn word embeddings from a corpus of 400 million tweets also used in---soria et al describe wordnet-lmf , an lmf model for representing wordnets which has been used in the ky-oto project | 0 |
word sense disambiguation ( wsd ) is the task of determining the correct meaning ( “ sense ” ) of a word in context , and several efforts have been made to develop automatic wsd systems---we use the glove algorithm to obtain 300-dimensional word embeddings from a union of these corpora | 0 |
neural machine translation has recently become the dominant approach to machine translation---in this paper , we propose a reinforcement learning based framework of dialogue system for automatic diagnosis | 0 |
this paper addresses the development and evaluation of pronunciation features for an automated system for scoring spontaneous speech---davidov and rappoport developed a framework which discovers concepts based on high frequency words and symmetry-based pattern graph properties | 0 |
to convert phrase trees to dependency structures , we followed the commonly used scheme---we converted the pcfg trees into dependency trees using the collins head rules | 1 |
the framework could be useful for machine translation applications and research in computational social science---yet effective method outperforms a number of baselines , and can be useful in translation applications and cross-cultural studies in computational social science | 1 |
the target language model was a trigram language model with modified kneser-ney smoothing trained on the english side of the bitext using the srilm tookit---the sri language modeling toolkit was used to train a trigram open-vocabulary language model with kneser-ney discounting on data that had boundary events inserted in the word stream | 1 |
through our experiments on japanese why-qa , we show that a combination of the above methods can improve why-qa accuracy---we use different pretrained word embeddings such as glove 1 and fasttext 2 as the initial word embeddings | 0 |
mcenery et al examined the distance of pronouns and their antecedent , and concluded that the antecedents of pronouns do exhibit clear patterns of distribution---mcenery et alexamined the distance of pronouns and their antecedent and concluded that the antecedents of pronouns do exhibit clear patterns of distribution | 1 |
in section 2 , we introduce and discuss the related work in this area---in section 2 , we introduce and discuss the related work | 1 |
we derive 100-dimensional word vectors using word2vec skip-gram model trained over the domain corpus---the skip-gram model adopts a neural network structure to derive the distributed representation of words from textual corpus | 1 |
the bleu score for all the methods is summarised in table 5---table 5 shows the bleu and per scores obtained by each system | 1 |
xie et al proposed to extract chinese abbreviations and their corresponding definitions based on anchor texts---the feature weights of the log-linear models were trained with the help of minimum error rate training and optimized for 4-gram bleu on the development test set | 0 |
argument mining ( am ) is a relatively new research area which involves , amongst others , the automatic detection in text of arguments , argument components , and relations between arguments ( see ( cite-p-10-1-13 ) for an overview )---argument mining is a trending research domain that focuses on the extraction of arguments and their relations from text | 1 |
this is because acquiring a large number of labeled data is expensive---a small number of labeled data can be enhanced by using additional unlabeled data | 1 |
to generate from the kbgen data , we induce a feature-based lexicalised tree adjoining grammar , augmented with a unification-based semantics from the training data---in particular , we use a feature-based lexicalized tree-adjoining grammar , that is derived from an hpsg grammar | 1 |
bengio et al presented a neural network language model where word embeddings are simultaneously learned along with a language model---we optimise the feature weights of the model with minimum error rate training against the bleu evaluation metric | 0 |
we used a 5-gram language model with modified kneser-ney smoothing , built with the srilm toolkit---we used the srilm toolkit to build unpruned 5-gram models using interpolated modified kneser-ney smoothing | 1 |
the obtained treebank is then transformed into ccg derivations---and then converts the trees into ccg derivations | 1 |
in this paper , we propose such a method , experimenting with semantic frame induction from linguistic and visual data---in this paper , we present a method to automatically learn argument role inventories for verbs | 1 |
we used a supervised learning approach with svm as the learning algorithm---c or , and the dso corpus , we trained supervised wsd systems with svm as the learning algorithm | 1 |
information extraction ( ie ) is the nlp field of research that is concerned with obtaining structured information from unstructured text---information extraction ( ie ) is a task of identifying 憽甪acts挕 ? ( entities , relations and events ) within unstructured documents , and converting them into structured representations ( e.g. , databases ) | 1 |
the proposed method will be incorporated into the tool kit for linguistic knowledge acquisition which we are now developing---barzilay and mckeown extract both singleand multiple-word paraphrases from a monolingual parallel corpus | 0 |
we use 300 dimension word2vec word embeddings for the experiments---we trained word embeddings using word2vec on 4 corpora of different sizes and types | 1 |
we follow a supervised approach , exploiting a svm polynomial kernel classifier trained with the challenge data---we have adopted a supervised approach , a svm polynomial kernel classifier trained with the data provided by the challenge | 1 |
berant et al introduced entailment graphs that provided a high-quality subsumption hierarchy---berant et al built a lexical entailment knowledge graph given the predicted results from the base classifier | 1 |
case-insensitive bleu-4 is our evaluation metric---the evaluation metric is case-sensitive bleu-4 | 1 |
sentence compression is the task of compressing long , verbose sentences into short , concise ones---sentence compression is the task of shortening a sentence while preserving its important information and grammaticality | 1 |
the structured approach also gives rise to a semi-supervised method , making it possible to take advantage of the readily available unlabeled data---nature of this problem gives rise to a better way of making use of the readily available unlabeled data , which further improves the proposed method | 1 |
due to their ability to capture syntactic and semantic information of words from large scale unlabeled texts , we pre-train the word embeddings from the given training dataset by word2vec toolkit---since our dataset is not so large , we make use of pre-trained word embeddings , which are trained on a much larger corpus with word2vec toolkit | 1 |
this enables the compositional operators to be learned by backpropagation from discourse annotations---compositional operators can be fine-tuned by backpropagating supervision from task-specific labels , enabling accurate and fast models | 1 |
we use the word2vec skip-gram model to train our word embeddings---we used 300 dimensional skip-gram word embeddings pre-trained on pubmed | 1 |
summarization is the process of condensing a source text into a shorter version while preserving its information content---summarization is a classic text processing problem | 1 |
language models were trained with the kenlm toolkit---we ’ ve demonstrated that the benefits of unsupervised multilingual learning increase steadily with the number of available languages | 0 |
etzioni et al present a system called knowitall , which implements an unsupervised domainindependent , bootstrapping approach to generate large facts of a specified ne from the web---etzioni et al presented the knowitall system that also utilizes hyponym patterns to extract class instances from the web | 1 |
the n-gram language models are trained using the srilm toolkit or similar software developed at hut---the lms are build using the srilm language modelling toolkit with modified kneserney discounting and interpolation | 1 |
coreference resolution is the task of determining when two textual mentions name the same individual---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 | 1 |
the labels were then transfered back into the target language---the labels were then transferred back into the target language | 1 |
brown et al present a hierarchical word clustering algorithm that can handle a large number of classes and a large vocabulary---brown et al use a very large amount of data , and a well-founded information theoretic model to induce large numbers of plausible semantic and syntactic clusters | 1 |
similarly to acd , our unconstrained system differed in that it also used word embeddings as features---as described in section 2 . 1 , our unconstrained acd system used an ensemble of two systems , one based on word embeddings and one based on features | 1 |
importantly , word embeddings have been effectively used for several nlp tasks , such as named entity recognition , machine translation and part-of-speech tagging---word embedding has shown promising results in variety of the nlp applications , such as named entity recognition , sentiment analysis and parsing | 1 |
for sentence segmentation and tokenization up to and including full morphological disambiguation for all languages , we rely on the udpipe---we trained the l1-regularized logistic regression classifier implemented in liblinear | 0 |
in this section , we briefly describe several other related challenges we are actively working on---in this section , we briefly describe several other related challenges | 1 |
zheng et al investigate chinese character embeddings for joint word segmentation and pos tagging---zheng et al investigated chinese character embeddings for chinese word segmentation and part-of-speech tagging | 1 |
we optimise the feature weights of the model with minimum error rate training against the bleu evaluation metric---we used minimum error rate training to optimize the feature weights | 0 |
the weights used during the reranking are tuned using the minimum error rate training algorithm---the feature weights are tuned to optimize bleu using the minimum error rate training algorithm | 1 |
in order to evaluate its quality versus the observed esl sentence , we use the meteor 2 and bleu evaluation metrics for machine translation---we use the automatic mt evaluation metrics bleu , meteor , and ter , to evaluate the absolute translation quality obtained | 1 |
their weights are optimized using minimum error-rate training on a held-out development set for each of the experiments---the weights 位 m in the log-linear model were trained using minimum error rate training with the news 2009 development set | 1 |
we perform pos tagging with the stanford pos tagger and create rules to switch the plurality of nouns---we use the stanford part-of-speech tagger and chunker to identify noun and verb phrases in the sentences | 1 |
in addition to more traditional components , such as knowledge-based and corpus-based metrics leveraged in a machine learning framework , we also use opinion analysis features to achieve a stronger semantic representation of textual units---to evaluate this type of similarity , we complement more traditional corpus and knowledge-based methods with opinion aware features , and use them in a meta-learning framework | 1 |
to obtain the vector representation of words , we used the google word2vec 1 , an open source tool---we used the penn treebank to perform empirical experiments on the proposed parsing models | 0 |
our parser produces a full syntactic parse of every sentence , and furthermore produces logical forms for portions of the sentence that have a semantic representation within the parser ’ s predicate vocabulary---parser produces a full syntactic parse of any sentence , while simultaneously producing logical forms for portions of the sentence that have a semantic representation within the parser ’ s predicate vocabulary | 1 |
in addition , we use an english corpus of roughly 227 million words to build a target-side 5-gram language model with srilm in combination with kenlm---the rule-based classifier of uchiyama et al incorporates syntactic information about japanese compound verbs , a type of mwe composed of two verbs | 0 |
the word embeddings were built from 200 million tweets using the word2vec model---their word embeddings were generated with word2vec , and trained on the arabic gigaword corpus | 1 |
we use several classifiers including logistic regression , random forest and adaboost implemented in scikit-learn---relation extraction is the task of predicting semantic relations over entities expressed in structured or semi-structured text | 0 |
distributional semantic models extract vectors representing word meaning by relying on the distributional hypothesis , that is , the idea that words that are related in meaning will tend to occur in similar contexts---distributional semantic models build on the distributional hypothesis which states that the meaning of a word can be modelled by observing the contexts in which it is used | 1 |
we use a random forest classifier for all experiments---we use a random forest classifier , as implemented in scikit-learn | 1 |
we estimate a 5-gram language model using interpolated kneser-ney discounting with srilm---we trained a 4-gram language model on this data with kneser-ney discounting using srilm | 1 |
the translation quality is evaluated by case-insensitive bleu and ter metric---we initialize our word vectors with 300-dimensional word2vec word embeddings | 0 |
otherwise , translations with wrong word order often lead to misunderstanding and incomprehensibility---lda is a generative probabilistic model where documents are viewed as mixtures over underlying topics , and each topic is a distribution over words | 0 |
sentences are passed through the stanford dependency parser to identify the dependency relations---the phrase structure trees produced by the parser are further processed with the stanford conversion tool to create dependency graphs | 1 |
for training on the training set , their results are : 0.69 f1 overall and 0.72 f1 for subtask b on---for training on the training set , their results are : 0 . 69 f1 overall and 0 . 72 f1 | 1 |
grosz and sidner claim that discourse segmentation is an important factor , though obviously not the only one , governing the use of referring expressions---grosz and sidner argue that such relations between intentions are a crucial part of intentional structure | 1 |
to further enhance the model performance , we use byte pair encoding with a coding size of 40k to segment the sentences of the training data into subwords---in order to limit the size of the vocabulary of the unmt model , we segmented tokens in the training data into sub-word units via byte pair encoding | 1 |
chen et al proposed a gated recursive neural network to incorporate context information---while oxford-style debates are a particularly convenient setting for studying the effects of conversational flow | 0 |
in this paper , we propose two new techniques to improve the current result---in this work , we followed the supervised approach and proposed two novel techniques to improve the current | 1 |
the standard classifiers are implemented with scikit-learn---for these experiments we use a maximum entropy classifier using the liblinear toolkit 2 | 0 |
the underlying parsing model is the dependency model with valance---the models we use are based on the generative dependency model with valence | 1 |
for relation classification , socher et al proposed a recursive matrix-vector model based on constituency parse trees---luong et al utilized the morpheme segments produced by morfessor and constructed morpheme trees for words to learn morphologically-aware word embeddings by the recursive neural network | 0 |
we use several classifiers including logistic regression , random forest and adaboost implemented in scikit-learn---we used the logistic regression implemented in the scikit-learn library with the default settings | 1 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.