Belief Revision and Argumentation Theory

MARCELO A. FALAPPA; GABRIELE KERN-ISBERNER; GUILLERMO R. SIMARI

Argumentation in Artificial Intelligence

Springer

Lugar: New York, United States of America; Año: 2009; p. 341 - 360

Belief revision is the process of changing beliefs to adapt the epistemic state of an agent to a new piece of information. The logical formalization of belief revision is a topic of research in philosophy, logic, and in computer science, in areas such as databases or artificial intelligence. On  the other hand, argumentation is concerned primarily with the evaluation of claims based on premises in order to reach conclusions. Both provide basic and substantial techniques for the art of reasoning, as it is performed by human beings in everyday life situations and which goes far beyond logical deduction. Reasoning, in this sense, makes possible to deal successfully with problems in uncertain, dynamic environments and has been promoting the development of human societies. The interest of computer scientists in both domains has increased considerably over the  past years, as agent systems are to be endowed with similar capabilities. In an agent environment,  belief revision describes the way in which an agent is supposed to change her beliefs when new information arrives, or changes in the world are observed; argumentation deals with strategies agents employ for their own reasoning, or to change the beliefs of other agents, by  providing reasons for such change. In this chapter, we will elaborate on the relationships between argumentation and belief revision, first recalling important work done by others and ourselves concerning the links between both areas. Based on such insights, we will develop a conceptual view on this topic, which is based on the understanding of argumentation and belief revision being complementary disciplines for the broad picture sketched above. Each needs the other?s support if we want to model successful decision making in a real world application. We will also discuss how one area may contribute to the other, enriching the respective framework.