Ranking kernels for structures and embeddings: A hybrid preference and classification model

Kateryna Tymoshenko, Daniele Bonadiman, Alessandro Moschitti

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Recent work has shown that Tree Kernels (TKs) and Convolutional Neural Networks (CNNs) obtain the state of the art in answer sentence reranking. Additionally, their combination used in Support Vector Machines (SVMs) is promising as it can exploit both the syntactic patterns captured by TKs and the embeddings learned by CNNs. However, the embeddings are constructed according to a classification function, which is not directly exploitable in the preference ranking algorithm of SVMs. In this work, we propose a new hybrid approach combining preference ranking applied to TKs and pointwise ranking applied to CNNs. We show that our approach produces better results on two well-known and rather different datasets: WikiQA for answer sentence selection and SemEval cQA for comment selection in Community Question Answering.

Original languageEnglish
Title of host publicationEMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings
PublisherAssociation for Computational Linguistics (ACL)
Pages897-902
Number of pages6
ISBN (Electronic)9781945626838
Publication statusPublished - 1 Jan 2017
Event2017 Conference on Empirical Methods in Natural Language Processing, EMNLP 2017 - Copenhagen, Denmark
Duration: 9 Sep 201711 Sep 2017

Publication series

NameEMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings

Conference

Conference2017 Conference on Empirical Methods in Natural Language Processing, EMNLP 2017
CountryDenmark
CityCopenhagen
Period9/9/1711/9/17

Fingerprint

Neural networks
Support vector machines
Syntactics

ASJC Scopus subject areas

  • Computer Science Applications
  • Information Systems
  • Computational Theory and Mathematics

Cite this

Tymoshenko, K., Bonadiman, D., & Moschitti, A. (2017). Ranking kernels for structures and embeddings: A hybrid preference and classification model. In EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings (pp. 897-902). (EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings). Association for Computational Linguistics (ACL).

Ranking kernels for structures and embeddings : A hybrid preference and classification model. / Tymoshenko, Kateryna; Bonadiman, Daniele; Moschitti, Alessandro.

EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings. Association for Computational Linguistics (ACL), 2017. p. 897-902 (EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Tymoshenko, K, Bonadiman, D & Moschitti, A 2017, Ranking kernels for structures and embeddings: A hybrid preference and classification model. in EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings. EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings, Association for Computational Linguistics (ACL), pp. 897-902, 2017 Conference on Empirical Methods in Natural Language Processing, EMNLP 2017, Copenhagen, Denmark, 9/9/17.
Tymoshenko K, Bonadiman D, Moschitti A. Ranking kernels for structures and embeddings: A hybrid preference and classification model. In EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings. Association for Computational Linguistics (ACL). 2017. p. 897-902. (EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings).
Tymoshenko, Kateryna ; Bonadiman, Daniele ; Moschitti, Alessandro. / Ranking kernels for structures and embeddings : A hybrid preference and classification model. EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings. Association for Computational Linguistics (ACL), 2017. pp. 897-902 (EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings).
@inproceedings{4aa5091a204e4090a212196d16f096ed,
title = "Ranking kernels for structures and embeddings: A hybrid preference and classification model",
abstract = "Recent work has shown that Tree Kernels (TKs) and Convolutional Neural Networks (CNNs) obtain the state of the art in answer sentence reranking. Additionally, their combination used in Support Vector Machines (SVMs) is promising as it can exploit both the syntactic patterns captured by TKs and the embeddings learned by CNNs. However, the embeddings are constructed according to a classification function, which is not directly exploitable in the preference ranking algorithm of SVMs. In this work, we propose a new hybrid approach combining preference ranking applied to TKs and pointwise ranking applied to CNNs. We show that our approach produces better results on two well-known and rather different datasets: WikiQA for answer sentence selection and SemEval cQA for comment selection in Community Question Answering.",
author = "Kateryna Tymoshenko and Daniele Bonadiman and Alessandro Moschitti",
year = "2017",
month = "1",
day = "1",
language = "English",
series = "EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings",
publisher = "Association for Computational Linguistics (ACL)",
pages = "897--902",
booktitle = "EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings",

}

TY - GEN

T1 - Ranking kernels for structures and embeddings

T2 - A hybrid preference and classification model

AU - Tymoshenko, Kateryna

AU - Bonadiman, Daniele

AU - Moschitti, Alessandro

PY - 2017/1/1

Y1 - 2017/1/1

N2 - Recent work has shown that Tree Kernels (TKs) and Convolutional Neural Networks (CNNs) obtain the state of the art in answer sentence reranking. Additionally, their combination used in Support Vector Machines (SVMs) is promising as it can exploit both the syntactic patterns captured by TKs and the embeddings learned by CNNs. However, the embeddings are constructed according to a classification function, which is not directly exploitable in the preference ranking algorithm of SVMs. In this work, we propose a new hybrid approach combining preference ranking applied to TKs and pointwise ranking applied to CNNs. We show that our approach produces better results on two well-known and rather different datasets: WikiQA for answer sentence selection and SemEval cQA for comment selection in Community Question Answering.

AB - Recent work has shown that Tree Kernels (TKs) and Convolutional Neural Networks (CNNs) obtain the state of the art in answer sentence reranking. Additionally, their combination used in Support Vector Machines (SVMs) is promising as it can exploit both the syntactic patterns captured by TKs and the embeddings learned by CNNs. However, the embeddings are constructed according to a classification function, which is not directly exploitable in the preference ranking algorithm of SVMs. In this work, we propose a new hybrid approach combining preference ranking applied to TKs and pointwise ranking applied to CNNs. We show that our approach produces better results on two well-known and rather different datasets: WikiQA for answer sentence selection and SemEval cQA for comment selection in Community Question Answering.

UR - http://www.scopus.com/inward/record.url?scp=85051568880&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85051568880&partnerID=8YFLogxK

M3 - Conference contribution

T3 - EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings

SP - 897

EP - 902

BT - EMNLP 2017 - Conference on Empirical Methods in Natural Language Processing, Proceedings

PB - Association for Computational Linguistics (ACL)

ER -