Abstract
This paper describes a method for knowledge sharing in a community using associative representation. We propose two ideas. One is associative representation for facilitating externalization of both personal and community information. The associative representation links heterogenenous information without definding the semantics strictly. We leave the interpretation of the semantics to human tacit knowledge. The other is visualization of information interaction using talking-agents metaphor. Taking-agents metaphor mimics a salon in which agents representing each community member interact with each others, thereby the user can see how their own or others' knowledge interact. We have developed a system called CoMeMo-Community that pursues collaborative story generation based on the talking-agents metaphor. We investigated how far people can exchange ideas with associative representation and how people react the talking-agents metaphor.
