A collaborative filtering system using clustering and genetic algorithms

Abstract

Recommender systems have been essential these days to assist online customers to acquire useful information. However, one of the popular types of the systems called memory-based collaborative filtering suffers from several fundamental problems in spite of its main advantages such as simplicity and efficiency. This study addresses the scalability problem which is one of major problems of the system. We employ a clustering technique to handle the problem and propose a novel idea using the genetic algorithm to enhance the performance of the system in terms of prediction accuracy, not to mention scalability. Experimental results demonstrated successful performance achievements of the proposed method under various data conditions.