text stringlengths 82 736 | label int64 0 1 |
|---|---|
here we present a series of experiments that led us to this conclusion---paper will describe a series of carefully-designed experiments that led us to these conclusions | 1 |
we investigate active learning methods for japanese dependency parsing---we describe our proposed methods and others of active learning for japanese dependency parsing | 1 |
the training set is very small , and it is a known fact that generative models tend to work better for small datasets and discriminative models tend to work better for larger datasets---we can use both subtree-and cluster-based features for parsing models | 0 |
this approach yielded a precision between 71 % and 82 % on the news headline dataset---evaluated on a news headline dataset , our model yielded higher accuracy | 1 |
we use the stanford part-of-speech tagger and chunker to identify noun and verb phrases in the sentences---we use stanford part-of-speech tagger to automatically detect nouns from text | 1 |
our baseline system is a popular phrase-based smt system , moses , with 5-gram srilm language model , tuned with minimum error training---table 2 shows the inter-annotator agreement of analytic scores for each prompt in kappa and qwk | 0 |
we learn a distance metric for each category node , and measure entity vector similarity under aggregated metrics---such that we measures entity-context similarity under aggregated distance metrics of hierarchical category nodes | 1 |
discourse parsing is a fundamental task in natural language processing that entails the discovery of the latent relational structure in a multi-sentence piece of text---the n-gram models are created using the srilm toolkit with good-turning smoothing for both the chinese and english data | 0 |
we train a kn-smoothed 5-gram language model on the target side of the parallel training data with srilm---we created 5-gram language models for every domain using srilm with improved kneserney smoothing on the target side of the training parallel corpora | 1 |
ahmad and kondrak , 2005 ) proposed a spelling error model from search query logs to improve the quality of query---an important aspect of simplification is syntactic transformation in which sentences deemed difficult are re-written as multiple sentences | 0 |
benefitting from the hierarchical semantic knowledge , the proposed approach alleviates the overfitting risk in a knowledge-driven manner---in a data-driven manner , this study introduces semantic knowledge into the splitting process | 1 |
we calculate lexical surprisal of each word in our corpus by training a simple trigram model over words on the open american national corpus using the srilm toolkit---to calculate language model features , we train traditional n-gram language models with ngram lengths of four and five using the srilm toolkit | 1 |
we used the penn treebank wsj corpus to perform empirical experiments on the proposed parsing models---we used stanford corenlp to tokenize the english and german data according to the penn treebank standard | 1 |
we trained two 5-gram language models on the entire target side of the parallel data , with srilm---we used srilm to build a 4-gram language model with kneser-ney discounting | 1 |
below , we review the orthogonal parameters of segmentation , segment order and segment contiguity ( § 2 )---as described herein , for use with mt systems , we propose a new automatic evaluation method | 0 |
and while discourse parsing is a document level task , discourse segmentation is done at the sentence level , assuming that sentence boundaries are known---discourse parsing is a challenging task and plays a critical role in discourse analysis | 1 |
methods for fine-grained sentiment analysis are developed by hu and liu , ding et al and popescu and etzioni---in their model , citing articles ¡° vote ¡± on each cited article ¡¯ s topic distribution | 0 |
we use word2vec , with the parameters suggested in the udpipe manual---we use the word2vec tool to pre-train the word embeddings | 1 |
in particular , we use a generalized version of mira that can incorporate k-best decoding in the update procedure---in our model , we apply a generalized version of mira that can incorporate k-best decoding in the update procedure | 1 |
the gro task concerns the automatic annotation of documents with gene regulation ontology concepts---the gro task aims to populate the gene regulation ontology with events and relations identified from text | 1 |
the reference corpora and data sets are pos tagged with the ims treetagger---the pos tags used in the reordering model are obtained using the treetagger | 1 |
in this study , engagement is considered as a sentiment as to whether users like intelligent assistants and feel like they want to use them continually---in their studies , user satisfaction was measured as to whether intelligent assistants can accomplish predefined tasks | 1 |
we use the l2-regularized logistic regression of liblinear as our term candidate classifier---the language models are 4-grams with modified kneser-ney smoothing which have been trained with the srilm toolkit | 0 |
in our experiments , we use the english-french part of the europarl corpus---we extract our paraphrase grammar from the french-english portion of the europarl corpus | 1 |
wiegand et al proposed to use high-level features by combining several linguistic features and lexicons of abusive words in the cross-domain classification of abusive microposts from different sources---wiegand et al used feature-based classification to build a lexicon of abusive words , which is similar to the interpretability task in this paper of identifying indicative unigram features | 1 |
what information to be included in a report---what information is to be included in a report | 1 |
we use moses toolkit for pbsmt training and sockeye toolkit for nmt training---we use the popular moses toolkit to build the smt system | 1 |
next , we present a flexible learning framework to learn distributed word representation based on the ordinal semantic knowledge---as our case study , our analyses of the hidden activation patterns show that the v isual model learns an abstract representation of the information structure of a single sentence in the language , and pays selective attention to lexical categories and grammatical functions that carry semantic information | 0 |
the re-ranking algorithms include rescoring and minimum bayes-risk decoding---this list is then rescored using minimum bayes-risk decoding | 1 |
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 )---we use the logistic regression classifier in the skll package , which is based on scikit-learn , optimizing for f 1 score | 0 |
in this paper , we integrate the cost from a graphbased model which directly models dependency links---in this paper , we exploit second-order relations , similar to the second-order edge factorization of dependency trees | 1 |
csc is a task in which each training instance has a vector of misclassification costs associated with it , thus rendering some mistakes to be more expensive than others---csc is a task in which each training instance has a vector of misclassification costs associated with it , thus rendering some mistakes on some instances to be more expensive than others | 1 |
to resolve the problem of generating a grammatically incorrect sentence , our method uses dependency structures and japanese dependency constraints to determine the word order of a translation---sentence , our method utilizes dependency structures and japanese dependency constraints to determine the word order of a translation | 1 |
the language model has an embedding size of 250 and two lstm layers with a hidden size of 1000---the decoder and encoder word embeddings are of size 620 , the encoder uses a bidirectional layer with 1000 lstms to encode the source side | 1 |
text mining results are presented as a browsable variable hierarchy which allows users to inspect all mentions of a particular variable type in the text as well as any generalisations or specialisations---text mining results are then presented as a browsable variable hierarchy which allows users to inspect all mentions of a particular variable type in the text | 1 |
the asymmetric alignments are symmetrized with the intersection and the grow-diag-final-and heuristics---after em , we obtain a symmetrized alignment by applying the grow-diag-final-and heuristic to the two trained alignments | 1 |
similarly , turian et al collectively used brown clusters , cw and hlbl embeddings , to improve the performance of named entity recognition and chucking tasks---for example , turian et al have improved the performance of chunking and named entity recognition by using word embedding also as one of the features in their crf model | 1 |
coster and kauchak extend a pbmt model to include phrase deletion and outperform coster and kauchak---coster and kauchak and wubben et al use a modified phrase-based model based on a machine translation framework | 1 |
for the task of event trigger prediction , we train a multi-class logistic regression classifier using liblinear---the target-side language models were estimated using the srilm toolkit | 0 |
word embeddings have proven to be effective models of semantic representation of words in various nlp tasks---the srilm toolkit was used to build this language model | 0 |
we implement the classifiers using the text classification framework dkpro tc which includes all of the abovementioned classifiers---mead is a centroid based multi document summarizer , which generates summaries using cluster centroids produced by topic detection and tracking system | 0 |
we further used a 5-gram language model trained using the srilm toolkit with modified kneser-ney smoothing---for the fluency and grammaticality features , we train 4-gram lms using the development dataset with the sri toolkit | 1 |
the smt system was tuned on the development set newstest10 with minimum error rate training using the bleu error rate measure as the optimization criterion---system tuning was carried out using minimum error rate training optimised with k-best mira on a held out development set | 1 |
we use a standard phrasebased translation system---we use the moses phrase-based mt system with standard features | 1 |
tjp was focused on the ‘ constrained ’ task , which used only training and development data provided---as ‘ constrained ’ , which used only the provided training and development data | 1 |
to encode the original sentences we used word2vec embeddings pre-trained on google news---we trained word embeddings using word2vec on 4 corpora of different sizes and types | 1 |
we use glove word embeddings , which are 50-dimension word vectors trained with a crawled large corpus with 840 billion tokens---we use the 200-dimensional global vectors , pre-trained on 2 billion tweets , covering over 27-billion tokens | 1 |
hypernym discovery aims to extract such noun pairs that one noun is a hypernym of the other---hypernym discovery is a task to extract such noun pairs that one noun is a hypernym of the other | 1 |
instead , we formalize dependency parsing as the problem of independently selecting the head of each word in a sentence---in our work , we formalize dependency parsing as the task of finding for each word in a sentence | 1 |
here , we extend the first approach , and show that better lexical generalization provides significant performance gains---in this paper , we demonstrate that significant performance gains can be achieved in ccg | 1 |
word sense disambiguation ( wsd ) is the task of identifying the correct sense of an ambiguous word in a given context---we compare our approach to the lcseg algorithm and use sentences as segmentation unit | 0 |
coreference resolution is a set partitioning problem in which each resulting partition refers to an 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 |
to tackle this issue , we leverage pretrained word embeddings , specifically the 300 dimension glove embeddings trained on 42b tokens of external text corpora---for the character-based model we use publicly available pre-trained character embeddings 3 de- rived from glove vectors trained on common crawl | 1 |
adapting our classifier to the task , we obtain 72.4 % accuracy , only 2.3 % below state of the art results---applying our story cloze classifier to this dataset yields 53 . 2 % classification accuracy | 1 |
we also examine similar classes in portuguese , and the predictive powers of alternations in this language with respect to the same semantic components---we also have begun to examine related classes in portuguese , and find that these verbs demonstrate similarly coherent | 1 |
semeval 2014 is a semantic evaluation of natural language processing ( nlp ) that comprises several tasks---semeval is the international workshop on semantic evaluation , formerly senseval | 1 |
we implement an in-domain language model using the sri language modeling toolkit---we train a trigram language model with the srilm toolkit | 1 |
sentiment analysis ( sa ) is the task of analysing opinions , sentiments or emotions expressed towards entities such as products , services , organisations , issues , and the various attributes of these entities ( cite-p-9-3-3 )---sentiment analysis ( sa ) is the task of determining the sentiment of a given piece of text | 1 |
we also want to make better use of the complex transition system to address the data sparsity issue for neural amr parsing---to address the sparsity issue of neural amr parsing , we feed feature embeddings from the transition state | 1 |
the parsing system has been implemented and has confirmed the feasibility of ottr approach to the modeling of these phenomena---parsing algorithm has been implemented and has confirmed the feasibility of our approach to the modeling of these phenomena | 1 |
gram language models were trained with lmplz---five-gram language models are trained using kenlm | 1 |
in recent years , numerous methods have been carefully studied for ner task , including hidden markov models , support vector machines and conditional random fields---the reranking parser of charniak and johnson was used to parse the bnc | 0 |
for example , cut can be used in the sense of “ cutting costs , ” which carries with it restrictions on instruments , locations , and so on that somewhat overlap with eliminate as in “ eliminating costs .---for example , cut can be used in the sense of “ cutting costs , ” which carries with it restrictions on instruments , locations , and so on that somewhat overlap with eliminate | 1 |
information extraction ( ie ) is a technology that can be applied to identifying both sources and targets of new hyperlinks---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 |
to implement svm algorithm , we have used the publicly available python based scikit-learn package---classification based methods are effective for this task | 0 |
in future work , we intend to build on the work reported in this paper in several ways---djuric et al used paragraph embeddings for detecting hate speech | 0 |
the parameters of the log-linear model are tuned by optimizing bleu on the development data using mert---that occurs in a visual environment , and is crucial for language acquisition , when much of the linguistic content refers to the visual surroundings of the child ( cite-p-11-3-0 ) | 0 |
we measure translation quality via the bleu score---experimental results show 79 % hit rate on manually annotated aspect | 0 |
our model processes over 1,700 english sentences per second , which is 30 times faster than the sparse-feature method---model processes over 1 , 700 english sentences per second , which is 30 times faster than the sparse-feature method | 1 |
domain dependence is a well-known issue for supervised nlp tasks such as framenet srl---domain dependence is a major problem for supervised nlp tasks such as framenet | 1 |
the main focus of the parser is on argument spans---with the main focus of the parser being on argument spans | 1 |
mem2seq is the first neural generative model that combines the multi-hop attention over memories with the idea of pointer network---mem2seq combines the multi-hop attention mechanism in endto-end memory networks with the idea of pointer networks | 1 |
previous work consistently reported that the wordbased translation models yielded better performance than the traditional methods for question retrieval---we have proposed an active reward learning model using gaussian process classification and an unsupervised neural network-based dialogue embedding to enable truly online policy learning | 0 |
in such a case , the end-user may prefer a concise summary of the ongoing discussion to save time---in such a case , the end-user may prefer a concise summary of the ongoing discussion | 1 |
we propose a framework to select and rank mandatory matching phrases ( mmp ) for question answering---we built a 5-gram language model on the english side of europarl and used the kneser-ney smoothing method and srilm as the language model toolkit | 0 |
we use conceptnet and coreference resolution as external knowledge---our knowledge acquisition method follows the scheme of conceptnet | 1 |
the task is to classify whether each comment is relevant to the question---we propose a neural architecture which learns a distributional semantic representation that leverage both document and sentence level information | 0 |
we use the skip-gram model , trained to predict context tags for each word---we adopt pretrained embeddings for word forms with the provided training data by word2vec | 1 |
we leverage latent dirichlet allocation for topic discovery and modeling in the reference source---we use the term-sentence matrix to train a simple generative topic model based on lda | 1 |
we evaluate our systems in terms of topic relevance , which is different from prior research---word sense disambiguation ( wsd ) is a key task in computational lexical semantics , inasmuch as it addresses the lexical ambiguity of text by making explicit the meaning of words occurring in a given context ( cite-p-18-3-10 ) | 0 |
we use minimum error rate training to tune the feature weights of hpb for maximum bleu score on the development set with serval groups of different start weights---we set the feature weights by optimizing the bleu score directly using minimum error rate training on the development set | 1 |
we use a sequential combination of a rule-based approach and machine learning to extract definitions---a combination of a rule-based approach and machine learning is a good way to extract definitions from texts | 1 |
we used the sri language modeling toolkit to train lms on our training data for each ilr level---we built a 5-gram language model from it with the sri language modeling toolkit | 1 |
this method employed ranking svm , the learning to rank method , to perform keyphrase extraction---to train a learning to rank model , ranking svm 9 , a powerful method for information retrieval , was adopted | 1 |
the feature weights for each system were tuned on development sets using the moses implementation of minimum error rate training---the ape system for each target language was tuned on comparable development sets , optimizing ter with minimum error rate training | 1 |
the remembrance agent is an early prototype of a continuously running automated information retrieval system , which was implemented as a plugin for the text editor emacs 3---in this section , we compare our work against other data-driven endto-end conversation | 0 |
for the actioneffect embedding model , we use pre-trained glove word embeddings as input to the lstm---for word-level embedding e w , we utilize pre-trained , 300-dimensional embedding vectors from glove 6b | 1 |
in this section , we describe the dirt algorithm for acquiring inference rules---this paper presents an entity-centric joint model for japanese | 0 |
reading comprehension ( rc ) is a high-level task in natural language understanding that requires reading a document and answering questions about its content---reading comprehension ( rc ) is the ability to read text , process it , and understand its meaning.2 how to endow computers with this capacity has been an elusive challenge and a long-standing goal of artificial intelligence ( e.g. , ( cite-p-16-1-10 ) ) | 1 |
we have presented a framework for word alignment based on log-linear models between parallel texts---in this paper , we present a framework for word alignment based on log-linear models | 1 |
in addition , we build a 5-gram continuous space language model for french---we also want to use a continuous space language model in an nbest list rescoring step after decoding | 1 |
seq2seq based conversation modeling approaches have been proven to be able to generate response directly---neural networks based models like seq2seq architecture are proven to be effective to generate valid responses for a dialogue system | 1 |
hu et al proposes integration of constraints coming in the form of first order logic rules during training of nns---hu et al employed knowledge distillation to enhance various types of neural networks with declarative firstorder logic rules | 1 |
in addition to these two key indicators , we evaluated the translation quality using an automatic measure , namely bleu score---whilst , the parameters for the maximum entropy model are developed based on the minimum error rate training method | 0 |
this result is important as it may fundamentally change the current binary classification paradigm---this result is important as it may fundamentally change the way that many practical classification | 1 |
we employed the machine learning tool of scikit-learn 3 , for training the classifier---we used the scikit-learn implementation of svrs and the skll toolkit | 1 |
to measure translation accuracy , we use the automatic evaluation measures of bleu and ribes measured over all sentences in the test corpus---in this paper , we have described a novel set of strategies for answering definition questions from multiple sources | 0 |
it has been shown that word embeddings are able to capture to certain semantic and syntactic aspects of words---extensive experiments have leveraged word embeddings to find general semantic relations | 1 |
in 25 of the sampled cases , at least one of the three systems made a change that improved the bleu score , whereas the score was adversely affected for at least one system in 13 cases---in 25 of the sampled cases , at least one of the three systems made a change that improved the bleu score , whereas the score was adversely affected for at least one system | 1 |
we used the naive bayes multinomial classifier and the alternating decision tree classifier from the weka toolkit---we applied the naive bayes probabilistic supervised learning algorithm from the weka machine learning library | 1 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.