A taxonomy of argumentation models used for knowledge representation
Citation: Jamal Bentahar, Bernard Moulin, Micheline Bélanger (2010) A taxonomy of argumentation models used for knowledge representation. Artificial Intelligence Review (RSS)
DOI (original publisher): 10.1007/s10462-010-9154-1
Semantic Scholar (metadata): 10.1007/s10462-010-9154-1
Sci-Hub (fulltext): 10.1007/s10462-010-9154-1
Internet Archive Scholar (search for fulltext): A taxonomy of argumentation models used for knowledge representation
Tagged: Computer Science
(RSS) argumentation (RSS), knowledge representation (RSS), artificial intelligence (RSS)
Summary
This paper proposes a taxonomy of argumentation models, distinguishing three main types of models, and comparing models in each of these categories:
- monological models - micro structure
- dialogical models - macro structure
- rhetorical models - audience's perception
Monological models
Monological models view arguments as a tentative proof, and focus on the internal structure of the chain of inference rules relating premises to conclusions
Dialogical models
Dialogical models emphasize the relationship between arguments. An argument can be seen as a dialogue game, where parties defend their viewpoint. In this view, argumentation is `defeasible' reasoning.
Rhetorical models
Rhetorical models study how arguments are used as a means of persuasion; they consider the audience's perception, and may relate to evaluative judgements (rather than truth).
Distinguishing between the models
Monological models are generally about the interal structure; Dialogical models are generally about the external structure. Rhetorical models are external to the argument, considering the communication aspects;
Joint models
- Both rhetorical and dialogical (Bench-Capon 2003; Bentahar et al. 2007b)
- Both monological and dialogical (Bench-Capon 1989; Farley and Freeman 1995; and Atkinson et al. model 2006)
Figure 1 summarizes the taxonomy, indicating the structure, foundation, and linkage of each type of model. The paper also presents an extensive description of various models, explaining the advantages and limits of each argumentation scheme considered.
Theoretical and Practical Relevance
Argumentation is an everyday human activity and computational argumentation is also widespread; this paper works towards developing a "global view of existing argumentation models and methods".
This is a seminal paper in argumentation which references and describes a large body of work, making sense of it with the taxonomy described.
The three types of models complement each other and should be combined.
