It takes three to tango: Triangulation approach to answer ranking in community question answering

Preslav Nakov, Lluis Marques, Francisco Guzmán

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

6 Citations (Scopus)

Abstract

We address the problem of answering new questions in community forums, by selecting suitable answers to already asked questions. We approach the task as an answer ranking problem, adopting a pairwise neural network architecture that selects which of two competing answers is better. We focus on the utility of the three types of similarities occurring in the triangle formed by the original question, the related question, and an answer to the related comment, which we call relevance, relatedness, and appropriateness. Our proposed neural network models the interactions among all input components using syntactic and semantic embeddings, lexical matching, and domain-specific features. It achieves state-of-the-art results, showing that the three similarities are important and need to be modeled together. Our experiments demonstrate that all feature types are relevant, but the most important ones are the lexical similarity features, the domain-specific features, and the syntactic and semantic embeddings.

Original languageEnglish
Title of host publicationEMNLP 2016 - Conference on Empirical Methods in Natural Language Processing, Proceedings
PublisherAssociation for Computational Linguistics (ACL)
Pages1586-1597
Number of pages12
ISBN (Electronic)9781945626258
Publication statusPublished - 1 Jan 2016
Event2016 Conference on Empirical Methods in Natural Language Processing, EMNLP 2016 - Austin, United States
Duration: 1 Nov 20165 Nov 2016

Publication series

NameEMNLP 2016 - Conference on Empirical Methods in Natural Language Processing, Proceedings

Conference

Conference2016 Conference on Empirical Methods in Natural Language Processing, EMNLP 2016
CountryUnited States
CityAustin
Period1/11/165/11/16

Fingerprint

Triangulation
Syntactics
Semantics
Neural networks
Network architecture
Experiments

ASJC Scopus subject areas

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

Cite this

Nakov, P., Marques, L., & Guzmán, F. (2016). It takes three to tango: Triangulation approach to answer ranking in community question answering. In EMNLP 2016 - Conference on Empirical Methods in Natural Language Processing, Proceedings (pp. 1586-1597). (EMNLP 2016 - Conference on Empirical Methods in Natural Language Processing, Proceedings). Association for Computational Linguistics (ACL).

It takes three to tango : Triangulation approach to answer ranking in community question answering. / Nakov, Preslav; Marques, Lluis; Guzmán, Francisco.

EMNLP 2016 - Conference on Empirical Methods in Natural Language Processing, Proceedings. Association for Computational Linguistics (ACL), 2016. p. 1586-1597 (EMNLP 2016 - Conference on Empirical Methods in Natural Language Processing, Proceedings).

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

Nakov, P, Marques, L & Guzmán, F 2016, It takes three to tango: Triangulation approach to answer ranking in community question answering. in EMNLP 2016 - Conference on Empirical Methods in Natural Language Processing, Proceedings. EMNLP 2016 - Conference on Empirical Methods in Natural Language Processing, Proceedings, Association for Computational Linguistics (ACL), pp. 1586-1597, 2016 Conference on Empirical Methods in Natural Language Processing, EMNLP 2016, Austin, United States, 1/11/16.
Nakov P, Marques L, Guzmán F. It takes three to tango: Triangulation approach to answer ranking in community question answering. In EMNLP 2016 - Conference on Empirical Methods in Natural Language Processing, Proceedings. Association for Computational Linguistics (ACL). 2016. p. 1586-1597. (EMNLP 2016 - Conference on Empirical Methods in Natural Language Processing, Proceedings).
Nakov, Preslav ; Marques, Lluis ; Guzmán, Francisco. / It takes three to tango : Triangulation approach to answer ranking in community question answering. EMNLP 2016 - Conference on Empirical Methods in Natural Language Processing, Proceedings. Association for Computational Linguistics (ACL), 2016. pp. 1586-1597 (EMNLP 2016 - Conference on Empirical Methods in Natural Language Processing, Proceedings).
@inproceedings{2e0b780046874e7d9077dd88545b19d0,
title = "It takes three to tango: Triangulation approach to answer ranking in community question answering",
abstract = "We address the problem of answering new questions in community forums, by selecting suitable answers to already asked questions. We approach the task as an answer ranking problem, adopting a pairwise neural network architecture that selects which of two competing answers is better. We focus on the utility of the three types of similarities occurring in the triangle formed by the original question, the related question, and an answer to the related comment, which we call relevance, relatedness, and appropriateness. Our proposed neural network models the interactions among all input components using syntactic and semantic embeddings, lexical matching, and domain-specific features. It achieves state-of-the-art results, showing that the three similarities are important and need to be modeled together. Our experiments demonstrate that all feature types are relevant, but the most important ones are the lexical similarity features, the domain-specific features, and the syntactic and semantic embeddings.",
author = "Preslav Nakov and Lluis Marques and Francisco Guzm{\'a}n",
year = "2016",
month = "1",
day = "1",
language = "English",
series = "EMNLP 2016 - Conference on Empirical Methods in Natural Language Processing, Proceedings",
publisher = "Association for Computational Linguistics (ACL)",
pages = "1586--1597",
booktitle = "EMNLP 2016 - Conference on Empirical Methods in Natural Language Processing, Proceedings",

}

TY - GEN

T1 - It takes three to tango

T2 - Triangulation approach to answer ranking in community question answering

AU - Nakov, Preslav

AU - Marques, Lluis

AU - Guzmán, Francisco

PY - 2016/1/1

Y1 - 2016/1/1

N2 - We address the problem of answering new questions in community forums, by selecting suitable answers to already asked questions. We approach the task as an answer ranking problem, adopting a pairwise neural network architecture that selects which of two competing answers is better. We focus on the utility of the three types of similarities occurring in the triangle formed by the original question, the related question, and an answer to the related comment, which we call relevance, relatedness, and appropriateness. Our proposed neural network models the interactions among all input components using syntactic and semantic embeddings, lexical matching, and domain-specific features. It achieves state-of-the-art results, showing that the three similarities are important and need to be modeled together. Our experiments demonstrate that all feature types are relevant, but the most important ones are the lexical similarity features, the domain-specific features, and the syntactic and semantic embeddings.

AB - We address the problem of answering new questions in community forums, by selecting suitable answers to already asked questions. We approach the task as an answer ranking problem, adopting a pairwise neural network architecture that selects which of two competing answers is better. We focus on the utility of the three types of similarities occurring in the triangle formed by the original question, the related question, and an answer to the related comment, which we call relevance, relatedness, and appropriateness. Our proposed neural network models the interactions among all input components using syntactic and semantic embeddings, lexical matching, and domain-specific features. It achieves state-of-the-art results, showing that the three similarities are important and need to be modeled together. Our experiments demonstrate that all feature types are relevant, but the most important ones are the lexical similarity features, the domain-specific features, and the syntactic and semantic embeddings.

UR - http://www.scopus.com/inward/record.url?scp=85029363661&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85029363661&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85029363661

T3 - EMNLP 2016 - Conference on Empirical Methods in Natural Language Processing, Proceedings

SP - 1586

EP - 1597

BT - EMNLP 2016 - Conference on Empirical Methods in Natural Language Processing, Proceedings

PB - Association for Computational Linguistics (ACL)

ER -