Towards a General Architecture for Building Intelligent, Flexible, and Adaptable Recommender System Based on MAS Technology

Nowadays, recommender systems have been used to reduce information overload and to find the items that are of interest to the user. Many techniques have been proposed for providing recommendations to consumers or users. All currently available recommender techniques have strengths and weaknesses. Thus, numerous researcher studies have attempted to develop techniques that would overcome the various limitations of current recommender systems by combining existing techniques in different ways. On the other hand, we have found that many currently available recommender systems are still designed for some restricted domains. This paper presents our attempt to use agent technology to enhance recommender systems based on agent’s property advantages with the goal to analyze and design a general architecture easily adaptable to several domains.