A behavior centric service discovery model for building knowledge advantage machine

Knowledge advantage Machine (KaM) is an advanced system for knowledge exploitation. In this paper, we propose a Behavior Centric Service Discovery model, which helps one or more knowledge-workers to discover useful knowledge objects dubbed JANs, and then link the JANs into a personal knowledge and a group knowledge network. We construct JAN as a service to semantically present three categories of service behavior: expected service behavior that presents what requestor expects it to serve; actual service behavior that presents how it offers its service; and quality evaluation that presents whether its service behavior is consistent with requestor's expectation by checking conformance. On the basis of JAN linking, we build a KaM discovery Agent prototype to implement JAN discovery and linking. This is illustrated using an academic research scenario. Experimental results show that the proposed method is feasible and effective.