Improving Recommendation by Exchanging Meta-Information

Interest-based recommendation (IBR) is a kind of knowledge based automated recommendation, in which agents exchange (meta-) information about their underlying goals using argumentation. This helps in improving the quantitative and qualitative utility of a recommendation. IBR combines hybrid recommender system with automated argumentation between agents. IBR also improves recommendation repair activity by discovering interesting alternatives based on user´s underlying mental attitude. This paper analyzes the role of interaction between agent´s goals to improve recommendation. We give an experimental analysis to show that with increase in knowledge transfer, the benefits of an interest-based recommendation also increase as compared to other recommendation technique without argumentation.