Multi-Agent Reputation Point System Framework

Abstract

An interview survey was conducted on respondents from service, manufacturing and education industries in Malaysia, to understand the processes of personal knowledge management (PKM) among knowledge workers. The findings show that personal knowledge network is enhanced when recommendations from associates outside the organisation are relied upon to identify the required knowledge experts. Thus the reputation of knowledge experts is known by some people in the network since it is the basis for assessing and deciding the reliability of the expertise required. This paper proposes a framework for a multi-agent system to search an existing network, analyse and manage reputation points in the process of identifying knowledge experts to fulfil the need of connecting to knowledge experts in managing personal knowledge. Recommendation on future work includes the technical possibility of expanding this multiagent system to be implemented in the Semantic Web.