This paper proposes a multi-agent architecture and a trust model with which to foster the reuse of information in organizations which use knowledge bases or knowledge management systems. The architecture and the model have been designed with the goal of giving support to communities of practices which are a means of sharing knowledge. However, members of these communities are currently often geographically distributed, and less trust therefore exists among members than in traditional co-localizated communities of practice. This situation has led us to propose our trust model, which can be used to calculate what piece of knowledge is more trustworthy. The architecture’s artificial agents will use this model to recommend the most appropriate knowledge to the community’s members.
Top2. Communities Of Practice
Intellectual capital and knowledge management are currently growing since knowledge is a critical factor for an organization’s competitive advantage (Kautz, 2004)(Kautz, 2004). This growth determines organizations’ performance by studying how well they manage their most critical knowledge. However, to manage this critical knowledge it has to be known what knowledge is, and although there is no consensus about a knowledge concept (Kakabadse, et al, 2001)(Kakabadse, 2001), there are several definitions of knowledge as in (Ackoff, 1989)(Ackoff, 1989) and (Davenport et al, 1998)(Davenport 1998). In our case, knowledge is going to be understood as in (Ackoff, 1989)(Ackoff, 1989), that is, as an appropriate collection of information, such that its intent is to be useful. In order to manage knowledge an important instrument are communities (Gebert et al., 2004; Malhotra, 2000)(Gebert et al., 2004), (Malhotra, 2000). A community can be defined as a group of socially interacting persons who are mutually tied to one another and regularly meet at a common place (Hillery, 1955)(Hillery, 1955). The development of Internet and groupware technologies led to a new kind of community “virtual communities” where members can or not meet one another face to face and they may exchange words and ideas through the mediation of computers networks (Geib et al., 2004)(Geib et al., 2004).