Argumentation in Artificial Intelligence
This volume is a systematic, expansive presentation of the major achievements in the intersection between two fields of inquiry: Argumentation Theory and Artificial Intelligence. Contributions from international researchers who have helped shape this dyna
- PDF / 545,436 Bytes
- 20 Pages / 439.37 x 666.142 pts Page_size
- 112 Downloads / 260 Views
Dialogue Games for Agent Argumentation Peter McBurney and Simon Parsons
1 Introduction The rise of the Internet and the growth of distributed computing have led to a major paradigm shift in software engineering and computer science. Until recently, the notion of computation has been variously construed as numerical calculation, as information processing, or as intelligent symbol analysis, but increasingly, it is now viewed as distributed cognition and interaction between intelligent entities [60]. This new view has major implications for the conceptualization, design, engineering and control of software systems, most profoundly expressed in the concept of systems of intelligent software agents, or multi-agent systems [99]. Agents are software entities with control over their own execution; the design of such agents, and of multi-agent systems of them, presents major research and software engineering challenges to computer scientists. One key challenge is the design of means of communication between intelligent agents. Considerable research effort has been expended on the design of artificial languages for agent communications, such as DARPA’s Knowledge Query and Manipulation Language (KQML) [33] and the Foundation for Intelligent Physical Agents’ (now IEEE FIPA) Agent Communications Language (FIPA ACL) [35]. These languages, and languages like them, have been designed to be widely applicable. As well as being a strength, this feature can also be a weakness: agents participating in conversations have too many choices of what to utter at each turn, and thus agent dialogues may endure a state-space explosion. Allowing sufficient flexibility of expression while avoiding state-space explosion had led agent communications researchers to the study of formal dialogue games; these are rule-governed interactions between two or more players (or agents), where
Department of Computer Science, University of Liverpool, UK e-mail: [email protected] Department of Computer and Information Science, Brooklyn College, New York, USA e-mail: [email protected]
I. Rahwan, G. R. Simari (eds.), Argumentation in Artificial Intelligence, c Springer Science+Business Media, LLC 2009 DOI 10.1007/978-0-387-98197-0 13,
261
262
McBurney and Parsons
each player “moves” by making utterances, according to a defined set of rules. Although their study dates to at least the time of Aristotle [5], dialogue games have found recent application in philosophy, computational linguistics and Artificial Intelligence (AI). In philosophy, dialogue games have been used to study fallacious reasoning [41, 62] and to develop a game-theoretic semantics for various logics, e.g., intuitionistic and for classical logics [59]. In linguistics, they have been used to explain sequences of human utterances [57], with subsequent application to machinebased natural language processing and generation [49], and to human-computer interaction [9]. Within computer science and AI, they have been applied to modeling complex human reasoning, for example in le
Data Loading...
