Architecture for Context-Aware Pro-Active Recommender System

Abstract

This paper presents an architecture of a context- aware pro-active recommen der system. The system uses contextual information in order to provide recommendations that are more suitable to the particular individual user. Reduction-based theory has been used in order to be able to use the contextual information besides the user and item components of traditional two dimensional recommender systems. The proposed recommender system provides recommendations pro-actively by using multi-agent technology. The inference engine of the system uses conditional probability and multi- attribute theory in the decision making of what recommendations to be provided to users.