Classifying types of ethos support and attack

Rory Duthie, Katarzyna Budzynska

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

21 Downloads (Pure)

Abstract

Endorsing the character of allies and destroying credibility of opponents is a powerful tactic for persuading others, impacting how we see politicians and how we vote in elections, for example. Our previous work demonstrated that ethos supports and attacks use different language, we hypothesise that further distinctions should be made in order to better understand and implement ethotic strategies which people use in real-life communication. In this paper, we use the Aristotelian concept of elements of ethos: practical wisdom, moral virtue and goodwill, to determine specific grounds on which speakers can be endorsed and criticised. We propose a classification of types of ethos supports and attacks which is empirically derived from our corpus. The manual classification obtains a reliable Cohen's kappa κ = 0.52 and weighted κ = 0.7. Finally, we develop a pipeline to classify ethos supports and attacks into their types depending on whether endorsement or criticism is grounded in wisdom, virtue or goodwill. The automatic classification obtains a solid improvement of macro-averaged F1-score over the baseline of 10%, 25%, 9% for one vs all classification, and 16%, 18%, 10% for pairwise classification.

Original languageEnglish
Title of host publicationComputational Models of Argument - Proceedings of COMMA 2018
EditorsSanjay Modgil, Katarzyna Budzynska, John Lawrence, Katarzyna Budzynska
PublisherIOS Press
Pages161-168
Number of pages8
Volume305
ISBN (Print)9781614999058
DOIs
Publication statusPublished - 2018
Event7th International Conference on Computational Models of Argument, COMMA 2018 - Warsaw, Poland
Duration: 12 Sep 201814 Sep 2018

Publication series

NameFrontiers in Artificial Intelligence and Applications
Volume305
ISSN (Print)0922-6389

Conference

Conference7th International Conference on Computational Models of Argument, COMMA 2018
CountryPoland
CityWarsaw
Period12/09/1814/09/18

Fingerprint

Macros
Pipelines
Communication

Keywords

  • Corpus Analysis
  • Elements of Ethos
  • Ethos Mining
  • Ethos-Logos
  • Wisdom-Virtue-Goodwill

Cite this

Duthie, R., & Budzynska, K. (2018). Classifying types of ethos support and attack. In S. Modgil, K. Budzynska, J. Lawrence, & K. Budzynska (Eds.), Computational Models of Argument - Proceedings of COMMA 2018 (Vol. 305, pp. 161-168). (Frontiers in Artificial Intelligence and Applications; Vol. 305). IOS Press. https://doi.org/10.3233/978-1-61499-906-5-161
Duthie, Rory ; Budzynska, Katarzyna. / Classifying types of ethos support and attack. Computational Models of Argument - Proceedings of COMMA 2018. editor / Sanjay Modgil ; Katarzyna Budzynska ; John Lawrence ; Katarzyna Budzynska. Vol. 305 IOS Press, 2018. pp. 161-168 (Frontiers in Artificial Intelligence and Applications).
@inproceedings{b0ef7084bc724f568a3471b1b7e2204d,
title = "Classifying types of ethos support and attack",
abstract = "Endorsing the character of allies and destroying credibility of opponents is a powerful tactic for persuading others, impacting how we see politicians and how we vote in elections, for example. Our previous work demonstrated that ethos supports and attacks use different language, we hypothesise that further distinctions should be made in order to better understand and implement ethotic strategies which people use in real-life communication. In this paper, we use the Aristotelian concept of elements of ethos: practical wisdom, moral virtue and goodwill, to determine specific grounds on which speakers can be endorsed and criticised. We propose a classification of types of ethos supports and attacks which is empirically derived from our corpus. The manual classification obtains a reliable Cohen's kappa κ = 0.52 and weighted κ = 0.7. Finally, we develop a pipeline to classify ethos supports and attacks into their types depending on whether endorsement or criticism is grounded in wisdom, virtue or goodwill. The automatic classification obtains a solid improvement of macro-averaged F1-score over the baseline of 10{\%}, 25{\%}, 9{\%} for one vs all classification, and 16{\%}, 18{\%}, 10{\%} for pairwise classification.",
keywords = "Corpus Analysis, Elements of Ethos, Ethos Mining, Ethos-Logos, Wisdom-Virtue-Goodwill",
author = "Rory Duthie and Katarzyna Budzynska",
year = "2018",
doi = "10.3233/978-1-61499-906-5-161",
language = "English",
isbn = "9781614999058",
volume = "305",
series = "Frontiers in Artificial Intelligence and Applications",
publisher = "IOS Press",
pages = "161--168",
editor = "Sanjay Modgil and Katarzyna Budzynska and John Lawrence and Katarzyna Budzynska",
booktitle = "Computational Models of Argument - Proceedings of COMMA 2018",
address = "Netherlands",

}

Duthie, R & Budzynska, K 2018, Classifying types of ethos support and attack. in S Modgil, K Budzynska, J Lawrence & K Budzynska (eds), Computational Models of Argument - Proceedings of COMMA 2018. vol. 305, Frontiers in Artificial Intelligence and Applications, vol. 305, IOS Press, pp. 161-168, 7th International Conference on Computational Models of Argument, COMMA 2018, Warsaw, Poland, 12/09/18. https://doi.org/10.3233/978-1-61499-906-5-161

Classifying types of ethos support and attack. / Duthie, Rory; Budzynska, Katarzyna.

Computational Models of Argument - Proceedings of COMMA 2018. ed. / Sanjay Modgil; Katarzyna Budzynska; John Lawrence; Katarzyna Budzynska. Vol. 305 IOS Press, 2018. p. 161-168 (Frontiers in Artificial Intelligence and Applications; Vol. 305).

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

TY - GEN

T1 - Classifying types of ethos support and attack

AU - Duthie, Rory

AU - Budzynska, Katarzyna

PY - 2018

Y1 - 2018

N2 - Endorsing the character of allies and destroying credibility of opponents is a powerful tactic for persuading others, impacting how we see politicians and how we vote in elections, for example. Our previous work demonstrated that ethos supports and attacks use different language, we hypothesise that further distinctions should be made in order to better understand and implement ethotic strategies which people use in real-life communication. In this paper, we use the Aristotelian concept of elements of ethos: practical wisdom, moral virtue and goodwill, to determine specific grounds on which speakers can be endorsed and criticised. We propose a classification of types of ethos supports and attacks which is empirically derived from our corpus. The manual classification obtains a reliable Cohen's kappa κ = 0.52 and weighted κ = 0.7. Finally, we develop a pipeline to classify ethos supports and attacks into their types depending on whether endorsement or criticism is grounded in wisdom, virtue or goodwill. The automatic classification obtains a solid improvement of macro-averaged F1-score over the baseline of 10%, 25%, 9% for one vs all classification, and 16%, 18%, 10% for pairwise classification.

AB - Endorsing the character of allies and destroying credibility of opponents is a powerful tactic for persuading others, impacting how we see politicians and how we vote in elections, for example. Our previous work demonstrated that ethos supports and attacks use different language, we hypothesise that further distinctions should be made in order to better understand and implement ethotic strategies which people use in real-life communication. In this paper, we use the Aristotelian concept of elements of ethos: practical wisdom, moral virtue and goodwill, to determine specific grounds on which speakers can be endorsed and criticised. We propose a classification of types of ethos supports and attacks which is empirically derived from our corpus. The manual classification obtains a reliable Cohen's kappa κ = 0.52 and weighted κ = 0.7. Finally, we develop a pipeline to classify ethos supports and attacks into their types depending on whether endorsement or criticism is grounded in wisdom, virtue or goodwill. The automatic classification obtains a solid improvement of macro-averaged F1-score over the baseline of 10%, 25%, 9% for one vs all classification, and 16%, 18%, 10% for pairwise classification.

KW - Corpus Analysis

KW - Elements of Ethos

KW - Ethos Mining

KW - Ethos-Logos

KW - Wisdom-Virtue-Goodwill

UR - http://www.scopus.com/inward/record.url?scp=85053920931&partnerID=8YFLogxK

U2 - 10.3233/978-1-61499-906-5-161

DO - 10.3233/978-1-61499-906-5-161

M3 - Conference contribution

SN - 9781614999058

VL - 305

T3 - Frontiers in Artificial Intelligence and Applications

SP - 161

EP - 168

BT - Computational Models of Argument - Proceedings of COMMA 2018

A2 - Modgil, Sanjay

A2 - Budzynska, Katarzyna

A2 - Lawrence, John

A2 - Budzynska, Katarzyna

PB - IOS Press

ER -

Duthie R, Budzynska K. Classifying types of ethos support and attack. In Modgil S, Budzynska K, Lawrence J, Budzynska K, editors, Computational Models of Argument - Proceedings of COMMA 2018. Vol. 305. IOS Press. 2018. p. 161-168. (Frontiers in Artificial Intelligence and Applications). https://doi.org/10.3233/978-1-61499-906-5-161