Co-operation framework of case-based reasoning agents for automated product recommendation

Baykal M., Alhajj R., Polat F.

JOURNAL OF EXPERIMENTAL & THEORETICAL ARTIFICIAL INTELLIGENCE, vol.17, no.3, pp.201-220, 2005 (SCI-Expanded) identifier identifier


This paper proposes a co-operation framework for multiple role-based case-based reasoning (CBR) agents to handle the product recommendation problem for e-commerce applications. Each agent has different case structure with intersecting features and agents exploit all information related to the problem by co-operation, which is accomplished through the merge of distributed cases in order to form cases having better representation of the problem. The presented merge algorithm handles noisy distributed cases by negotiation on the difference values of the intersecting features. The role-based CBR agents merge the distributed cases by introducing a global heuristic function, which is used to evaluate the relevance of merged cases. The heuristic function exploits the relevancy of each merged case within the viewpoint of each agent and the satisfied/unsatisfied problem constraints. The viewpoint of an agent is represented by the value of consistency of distributed components of merged cases and agent s individual relevance values of the merged cases. Finally, the proposed framework has been tested for elective course recommendation.