Scrutable Feature Sets for Stance Classification

Angrosh Mandya, Advaith Siddharthan, Adam Wyner

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

Abstract

This paper describes and evaluates a novel feature set for stance classification of argumentative texts; i.e. deciding whether a post by a user is for or against the issue being debated.We model the debate both as attitude bearing features, including a set of automatically acquired ‘topic terms’ associated with a Distributional Lexical Model (DLM) that captures the writer’s attitude towards the topic term, and as dependency features that represent the points being made in the debate. The stance of the text towards the issue being debated is then learnt in a supervised framework as a function of these features. The main advantage of our feature set is that it is scrutable: The reasons for a classification can be explained to a human user in natural language. We also report that our method outperforms previous approaches to stance classification as well as a range of baselines based on sentiment analysis and topic-sentiment analysis.
Original languageEnglish
Title of host publicationProceedings of the 3rd Workshop on Argument Mining
PublisherACL Anthology
Pages60-69
Number of pages10
ISBN (Electronic)978-1-945626-17-3
Publication statusPublished - 2016
Event3rd Workshop on Argument Mining: ACL 2016 - Berlin, Germany
Duration: 12 Aug 201612 Aug 2016
http://argmining2016.arg.tech/ (3rd Workshop on Argument Mining - ACL 2016)

Conference

Conference3rd Workshop on Argument Mining
Abbreviated titleArgMining 2016
CountryGermany
CityBerlin
Period12/08/1612/08/16
Internet address

Fingerprint

Bearings (structural)

Cite this

Mandya, A., Siddharthan, A., & Wyner, A. (2016). Scrutable Feature Sets for Stance Classification. In Proceedings of the 3rd Workshop on Argument Mining (pp. 60-69). ACL Anthology.

Scrutable Feature Sets for Stance Classification. / Mandya, Angrosh; Siddharthan, Advaith; Wyner, Adam.

Proceedings of the 3rd Workshop on Argument Mining. ACL Anthology, 2016. p. 60-69.

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

Mandya, A, Siddharthan, A & Wyner, A 2016, Scrutable Feature Sets for Stance Classification. in Proceedings of the 3rd Workshop on Argument Mining. ACL Anthology, pp. 60-69, 3rd Workshop on Argument Mining, Berlin, Germany, 12/08/16.
Mandya A, Siddharthan A, Wyner A. Scrutable Feature Sets for Stance Classification. In Proceedings of the 3rd Workshop on Argument Mining. ACL Anthology. 2016. p. 60-69
Mandya, Angrosh ; Siddharthan, Advaith ; Wyner, Adam. / Scrutable Feature Sets for Stance Classification. Proceedings of the 3rd Workshop on Argument Mining. ACL Anthology, 2016. pp. 60-69
@inproceedings{15870391d3204aa4b2da64b18e6cdcbc,
title = "Scrutable Feature Sets for Stance Classification",
abstract = "This paper describes and evaluates a novel feature set for stance classification of argumentative texts; i.e. deciding whether a post by a user is for or against the issue being debated.We model the debate both as attitude bearing features, including a set of automatically acquired ‘topic terms’ associated with a Distributional Lexical Model (DLM) that captures the writer’s attitude towards the topic term, and as dependency features that represent the points being made in the debate. The stance of the text towards the issue being debated is then learnt in a supervised framework as a function of these features. The main advantage of our feature set is that it is scrutable: The reasons for a classification can be explained to a human user in natural language. We also report that our method outperforms previous approaches to stance classification as well as a range of baselines based on sentiment analysis and topic-sentiment analysis.",
author = "Angrosh Mandya and Advaith Siddharthan and Adam Wyner",
note = "This work was supported by the Economic and Social Research Council [Grant number ES/M001628/1].",
year = "2016",
language = "English",
pages = "60--69",
booktitle = "Proceedings of the 3rd Workshop on Argument Mining",
publisher = "ACL Anthology",

}

TY - GEN

T1 - Scrutable Feature Sets for Stance Classification

AU - Mandya, Angrosh

AU - Siddharthan, Advaith

AU - Wyner, Adam

N1 - This work was supported by the Economic and Social Research Council [Grant number ES/M001628/1].

PY - 2016

Y1 - 2016

N2 - This paper describes and evaluates a novel feature set for stance classification of argumentative texts; i.e. deciding whether a post by a user is for or against the issue being debated.We model the debate both as attitude bearing features, including a set of automatically acquired ‘topic terms’ associated with a Distributional Lexical Model (DLM) that captures the writer’s attitude towards the topic term, and as dependency features that represent the points being made in the debate. The stance of the text towards the issue being debated is then learnt in a supervised framework as a function of these features. The main advantage of our feature set is that it is scrutable: The reasons for a classification can be explained to a human user in natural language. We also report that our method outperforms previous approaches to stance classification as well as a range of baselines based on sentiment analysis and topic-sentiment analysis.

AB - This paper describes and evaluates a novel feature set for stance classification of argumentative texts; i.e. deciding whether a post by a user is for or against the issue being debated.We model the debate both as attitude bearing features, including a set of automatically acquired ‘topic terms’ associated with a Distributional Lexical Model (DLM) that captures the writer’s attitude towards the topic term, and as dependency features that represent the points being made in the debate. The stance of the text towards the issue being debated is then learnt in a supervised framework as a function of these features. The main advantage of our feature set is that it is scrutable: The reasons for a classification can be explained to a human user in natural language. We also report that our method outperforms previous approaches to stance classification as well as a range of baselines based on sentiment analysis and topic-sentiment analysis.

M3 - Conference contribution

SP - 60

EP - 69

BT - Proceedings of the 3rd Workshop on Argument Mining

PB - ACL Anthology

ER -