A decision tree for annotating argumentation scheme corpora

Research output: Contribution to conferencePaperpeer-review

39 Downloads (Pure)

Abstract

Quantitative approaches, necessary for e.g. computationallinguistic methods such as argument mining, require large annotated corpora of argumentative discourse. Publicly available corpora of argumentation schemes often only cover a small selection of example schemes and suffer from low
inter-annotator agreement. To address this, we present a heuristic decision tree for the classification of Walton's toplevel taxonomy of 60 schemes. An annotation study on 505 arguments resulted in a 97% classification covering 38 schemes (Cohen’s κ 0.723).
Original languageEnglish
Pages97-114
Publication statusPublished - 2020
Event3rd European Conference on Argumentation: Reason to Dissent - University of Groningen, Groningen, Netherlands
Duration: 24 Jun 201927 Feb 2021
http://ecargument.org

Conference

Conference3rd European Conference on Argumentation
CountryNetherlands
CityGroningen
Period24/06/1927/02/21
Internet address

Fingerprint Dive into the research topics of 'A decision tree for annotating argumentation scheme corpora'. Together they form a unique fingerprint.

Cite this