Identifying Argumentative Discourse Structures in Persuasive Essays

34828

Authors

nan

Abstract

In this paper, we present a novel ap- proach for identifying argumentative dis- course structures in persuasive essays. The structure of argumentation consists of sev- eral components (i.e. claims and premises) that are connected with argumentative re- lations. We consider this task in two consecutive steps. First, we identify the components of arguments using multiclass classification. Second, we classify a pair of argument components as either support or non-support for identifying the struc- ture of argumentative discourse. For both tasks, we evaluate several classifiers and propose novel feature sets including struc- tural, lexical, syntactic and contextual fea- tures. In our experiments, we obtain a macro F1-score of 0.726 for identifying argument components and 0.722 for argu- mentative relations. 1

Keywords

nan

Full Content

Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 46–56, October 25-29, 2014, Doha, Qatar. c⃝2014 Association for Computational Linguistics Identifying Argumentative Discourse Structures in Persuasive Essays Christian Stab† and Iryna Gurevych†‡ †Ubiquitous Knowledge Processing Lab (UKP-TUDA) Department of Computer Science, Technische Universit¨at Darmstadt ‡Ubiquitous Knowledge Processing Lab (UKP-DIPF) German Institute for Educational Research www.ukp.tu-darmstadt.de Abstract In this paper, we present a novel ap- proach for identifying argumentative dis- course structures in persuasive essays. The structure of argumentation consists of sev- eral components (i.e. claims and premises) that are connected with argumentative re- lations. We consider this task in two consecutive steps. First, we identify the components of arguments using multiclass classification. Second, we classify a pair of argument components as either support or non-support for identifying the struc- ture of argumentative discourse. For both tasks, we evaluate several classifiers and propose novel feature sets including struc- tural, lexical, syntactic and contextual fea- tures. In our experiments, we obtain a macro F1-score of 0.726 for identifying argument components and 0.722 for argu- mentative relations. 1 Introduction Argumentation is a crucial aspect of writing skills acquisition. The ability of formulating persuasive arguments is not only the foundation for convinc- ing an audience of novel ideas but also plays a ma- jor role in general decision making and analyzing different stances. However, current writing sup- port is limited to feedback about spelling, gram- mar, or stylistic properties and there is currently no system that provides feedback about written argu- mentation. By integrating argumentation mining in writing environments, students will be able to inspect their texts for plausibility and to improve the quality of their argumentation. An argument consists of several components. It includes a claim that is supported or attacked by at least one premise. The claim is the central compo- nent of an argument. It is a controversial statement that should not be accepted by the reader without additional support.1 The premise underpins the validity of the claim. It is a reason given by an author for persuading readers of the claim. Argu- mentative relations model the discourse structure of arguments. They indicate which argument com- ponents are related and constitute the structure of argumentative discourse. For example, the argu- ment in the following paragraph contains four ar- gument components: one claim (in bold face) and three premises (underlined). “(1) Museums and art galleries provide a better understanding about arts than Internet. (2) In most museums and art galleries, detailed descriptions in terms of the background, history and author are provided. (3) Seeing an artwork on- line is not the same as watching it with our own eyes, as (4) the picture online does not show the texture or three-di- mensional structure of the art, which is important to study.” In this example, the premises (2) and (3) sup- port the claim (1) whereas premise (4) is a support for premise (3). Thus, this example includes three argumentative support relations holding between the components (2,1), (3,1) …

Content truncated to 500 words

Labeled Text

In this paper, we present a novel approach for identifying argumentative discourse structures in persuasive essays. The structure of argumentation consists of several components (i.e. claims and premises) that are connected with argumentative relations. We consider this task in two consecutive steps. First, we identify the components of arguments using multiclass classification. Second, we classify a pair of argument components as either support or non-support for identifying the structure of argumentative discourse. For both tasks, we evaluate several classifiers and propose novel feature sets including structural, lexical, syntactic and contextual features. In our experiments, we obtain a macro F1-score of 0.726 for identifying argument components and 0.722 for argumentative relations. 1

Publication Details

Paper ID:
34828
Year:
2014
Publication:
TIPSTER
Publication Type:
conf
DOI:
nan