Task knowledge patterns reuse in multi-agent systems development

Search Swinburne Research Bank

Home

List of Titles

Please use this identifier to cite or link to this item: http://hdl.handle.net/1959.3/188743

Check Swinburne's subscriptions for full text availability

Title: Task knowledge patterns reuse in multi-agent systems development
Author(s): Cheah, WaiShiang; Sterling, Leon; Taveter, Kuldar
Abstract: Template-based knowledge models can be viewed as design patterns for specifying a task. The models can serve as reusable artifacts during the development of a multi agent system using the MAS-CommonKADS methodology. However, based on our observation of existing patterns, we note limitations of reusing those patterns in agent development. This paper presents task knowledge patterns that are described through our improved agent oriented template structure. The improved template structure presented in this paper provides an alternative approach to defining task knowledge patterns by incorporating a two dimensional view of agent oriented models. The task knowledge patterns introduced in this paper describe task knowledge in an agent context, while explicitly providing a description designed to encourage use and reuse in agent oriented software development. A demonstration of the reuse of task knowledge patterns in agent oriented modelling is presented in this paper. Specifically we show how a particular task knowledge pattern, selection of relevant source materials, can be used to rapidly prototype an adviser finder multi-agent system.
Publication type: Conference paper
Research centre: Swinburne University of Technology. Faculty of Information and Communication Technologies
Source: Lecture notes in computer science: Lecture notes in artificial intelligence: revised selected papers of the 13th International Conference on Principles and Practice of Multi-Agent Systems (PRIMA-2010), Kolkata, India, 12-15 November 2010 / Nirmit Desai, Alan Liu and Michael Winikoff (eds.), Vol. 7057, pp. 459-474
Publication year: 2012
Keyword(s): Agent-based systems; Multi-agent systems; Systems development
Publisher: Springer
ISSN: 0302-9743 (series ISSN)
ISBN: 9783642259197, 3642259197
Publisher URL: http://dx.doi.org/10.1007/978-3-642-25920-3_33
Peer reviewed