Filters








153 Hits in 2.7 sec

Hearst Patterns Revisited: Automatic Hypernym Detection from Large Text Corpora

Stephen Roller, Douwe Kiela, Maximilian Nickel
2018 Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)  
Methods for unsupervised hypernym detection may broadly be categorized according to two paradigms: pattern-based and distributional methods.  ...  Our results show that pattern-based models provide important contextual constraints which are not yet captured in distributional methods.  ...  We also thank Vered Shwartz, Enrico Santus, and Dominik Schlechtweg for providing us with their distributional spaces and baseline implementations.  ... 
doi:10.18653/v1/p18-2057 dblp:conf/acl/RollerKN18 fatcat:3r76wcbtu5ajvdvgo5aax2udji

Hypernyms under Siege: Linguistically-motivated Artillery for Hypernymy Detection [article]

Vered Shwartz, Enrico Santus, Dominik Schlechtweg
2017 arXiv   pre-print
Comparison to the state-of-the-art supervised methods shows that while supervised methods generally outperform the unsupervised ones, the former are sensitive to the distribution of training instances,  ...  Being based on general linguistic hypotheses and independent from training data, unsupervised measures are more robust, and therefore are still useful artillery for hypernymy detection.  ...  Unsupervised Hypernymy Detection Measures We experiment with a large number of unsupervised measures proposed in the literature for distributional hypernymy detection, with some new variants.  ... 
arXiv:1612.04460v2 fatcat:yhprqf6zebci3h2bchzdnswhxe

Hypernyms under Siege: Linguistically-motivated Artillery for Hypernymy Detection

Vered Shwartz, Enrico Santus, Dominik Schlechtweg
2017 Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers  
Comparison to the state-of-the-art supervised methods shows that while supervised methods generally outperform the unsupervised ones, the former are sensitive to the distribution of training instances,  ...  Being based on general linguistic hypotheses and independent from training data, unsupervised measures are more robust, and therefore are still useful artillery for hypernymy detection.  ...  Unsupervised Hypernymy Detection Measures We experiment with a large number of unsupervised measures proposed in the literature for distributional hypernymy detection, with some new variants.  ... 
doi:10.18653/v1/e17-1007 dblp:conf/eacl/SantusSS17 fatcat:gdtjxtmvm5clxb3qgddhziuwui

Improving Hypernymy Detection with an Integrated Path-based and Distributional Method

Vered Shwartz, Yoav Goldberg, Ido Dagan
2016 Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)  
Detecting hypernymy relations is a key task in NLP, which is addressed in the literature using two complementary approaches.  ...  Distributional methods, whose supervised variants are the current best performers, and path-based methods, which received less research attention.  ...  Distributional Methods Hypernymy detection is commonly addressed using distributional methods.  ... 
doi:10.18653/v1/p16-1226 dblp:conf/acl/ShwartzGD16 fatcat:yljga7p5t5borfg5bneh4s25bi

Improving Hypernymy Detection with an Integrated Path-based and Distributional Method [article]

Vered Shwartz, Yoav Goldberg, Ido Dagan
2016 arXiv   pre-print
Detecting hypernymy relations is a key task in NLP, which is addressed in the literature using two complementary approaches.  ...  Distributional methods, whose supervised variants are the current best performers, and path-based methods, which received less research attention.  ...  Distributional Methods Hypernymy detection is commonly addressed using distributional methods.  ... 
arXiv:1603.06076v3 fatcat:zsn34iw3unckzkwqwys4lf6zue

Relations such as Hypernymy: Identifying and Exploiting Hearst Patterns in Distributional Vectors for Lexical Entailment [article]

Stephen Roller, Katrin Erk
2016 arXiv   pre-print
We consider the task of predicting lexical entailment using distributional vectors.  ...  Both Diff and Asym can also be seen as a form of supervised Distributional Inclusion Hypothesis, with the vector difference being analogous to the set-inclusion measures of some unsupervised techniques  ...  Recently, other works have begun to analyze Concat and Diff for their ability to go beyond just hypernymy detection.  ... 
arXiv:1605.05433v2 fatcat:m4cip3ircza3tiu5mbtl3rszoe

Relations such as Hypernymy: Identifying and Exploiting Hearst Patterns in Distributional Vectors for Lexical Entailment

Stephen Roller, Katrin Erk
2016 Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing  
We consider the task of predicting lexical entailment using distributional vectors.  ...  Both Diff and Asym can also be seen as a form of supervised Distributional Inclusion Hypothesis, with the vector difference being analogous to the set-inclusion measures of some unsupervised techniques  ...  Recently, other works have begun to analyze Concat and Diff for their ability to go beyond just hypernymy detection.  ... 
doi:10.18653/v1/d16-1234 dblp:conf/emnlp/RollerE16 fatcat:bs2zecplenhpjh5ydb4boeyu2a

Pruning and repopulating a lexical taxonomy: experiments in Spanish, English and French

Rogelio Nazar, Antonio Balvet, Gabriela Ferraro, Rafael Marín, Irene Renau
2020 Journal of Intelligent Systems  
The first task is to identify and eliminate incorrect hypernymy links, and the second is to repopulate the taxonomy with new relations.  ...  In order to achieve these goals, we implemented a hybrid hypernym-detection algorithm that incorporates various intuitions, such as syntagmatic, paradigmatic and morphological association measures as well  ...  In [26] , there is invCL, as a measure that takes into account the distributional inclusion of x in y and the distributional non-inclusion of y in x.  ... 
doi:10.1515/jisys-2020-0044 fatcat:i53p6x63izevxf6ewkuzz5l5jm

Hypernym Detection Using Strict Partial Order Networks [article]

Sarthak Dash, Md Faisal Mahbub Chowdhury, Alfio Gliozzo, Nandana Mihindukulasooriya, Nicolas Rodolfo Fauceglia
2019 arXiv   pre-print
We apply it to induce hypernymy relations by training with is-a pairs.  ...  There also exist approaches (Weeds, Weir, and McCarthy 2004; Roller and Erk 2016; Shwartz, Santus, and Schlechtweg 2017) inspired by the Distributional Inclusion Hypothesis (DIH) (Geffet and Dagan 2005  ...  The first one is hypernym detection where the goal is to classify whether a given pair of terms are in a hypernymy relation.  ... 
arXiv:1909.10572v2 fatcat:rinjqpjecffgrc22oic6kswlhm

Hypernym Detection Using Strict Partial Order Networks

Sarthak Dash, Md Faisal Mahbub Chowdhury, Alfio Gliozzo, Nandana Mihindukulasooriya, Nicolas Rodolfo Fauceglia
2020 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
We apply it to induce hypernymy relations by training with is-a pairs.  ...  There also exist approaches (Weeds, Weir, and McCarthy 2004; Roller and Erk 2016; Shwartz, Santus, and Schlechtweg 2017) inspired by the Distributional Inclusion Hypothesis (DIH) (Geffet and Dagan 2005  ...  The first one is hypernym detection where the goal is to classify whether a given pair of terms are in a hypernymy relation.  ... 
doi:10.1609/aaai.v34i05.6263 fatcat:cj52zpiapbh6pem6cmobuiehyq

End-to-End Reinforcement Learning for Automatic Taxonomy Induction

Yuning Mao, Xiang Ren, Jiaming Shen, Xiaotao Gu, Jiawei Han
2018 Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)  
., detecting hypernymy pairs followed by organizing these pairs into a tree-structured hierarchy), we argue that such two-phase methods may suffer from error propagation, and cannot effectively optimize  ...  In our approach, the representations of term pairs are learned using multiple sources of information and used to determine which term to select and where to place it on the taxonomy via a policy network  ...  Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation hereon.  ... 
doi:10.18653/v1/p18-1229 dblp:conf/acl/HanRSMG18 fatcat:le3zugfx4vgfbg3ezqpvylnzt4

End-to-End Reinforcement Learning for Automatic Taxonomy Induction [article]

Yuning Mao, Xiang Ren, Jiaming Shen, Xiaotao Gu, Jiawei Han
2018 arXiv   pre-print
., detecting hypernymy pairs followed by organizing these pairs into a tree-structured hierarchy), we argue that such two-phase methods may suffer from error propagation, and cannot effectively optimize  ...  In our approach, the representations of term pairs are learned using multiple sources of information and used to determine which term to select and where to place it on the taxonomy via a policy network  ...  Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation hereon.  ... 
arXiv:1805.04044v1 fatcat:e7plggvgvrcgnnwhova2zq5bnm

Hearst Patterns Revisited: Automatic Hypernym Detection from Large Text Corpora [article]

Stephen Roller, Douwe Kiela, Maximilian Nickel
2018 arXiv   pre-print
Methods for unsupervised hypernym detection may broadly be categorized according to two paradigms: pattern-based and distributional methods.  ...  Our results show that pattern-based models provide important contextual constraints which are not yet captured in distributional methods.  ...  We also thank Vered Shwartz, Enrico Santus, and Dominik Schlechtweg for providing us with their distributional spaces and baseline implementations.  ... 
arXiv:1806.03191v1 fatcat:fzx4u5tforhenhyof6gkyngnpi

A Short Survey on Taxonomy Learning from Text Corpora: Issues, Resources and Recent Advances

Chengyu Wang, Xiaofeng He, Aoying Zhou
2017 Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing  
Roller et al. (2014) introduce the Selective Distributional Inclusion Hypothesis, which means the original DIH is correct, but only for relevant dimensions.  ...  Yet a further criticism is pointed out by Levy et al. (2015) .  ... 
doi:10.18653/v1/d17-1123 dblp:conf/emnlp/WangHZ17 fatcat:onmgzexqubbw7ixavokbrv2l64

Instantiation [article]

Abhijeet Gupta and Gemma Boleda and Sebastian Pado
2018 arXiv   pre-print
In computational linguistics, a large body of work exists on distributed modeling of lexical relations, focussing largely on lexical relations such as hypernymy (scientist -- person) that hold between  ...  We present a new, principled dataset for the task of instantiation detection as well as experiments and analyses on this dataset.  ...  Experiment 1: Instantiation Detection as Classification We frame instantiation detection as a supervised binary classification task.  ... 
arXiv:1808.01662v1 fatcat:qythiomfbrbklgokxmz7my2u2m
« Previous Showing results 1 — 15 out of 153 results