A collaborative filtering method based on associative memory model

Abstract

Recommender systems are intelligent systems that help consumers by recommending products they are likely to appreciate or purchase. These recommendations are based on the user's own purchasing, searching or browsing history and also that of other consumers with similar interests. These systems are often embedded in e-commerce applications with the aim to provide efficient personalized recommendations that are of mutual value to both the buyer and the seller. This paper presents a novel neural network based approach that employs associative memory model to make recommendations for purchase to consumers. Associative memory models are inherently able to solve pattern completion problem. This intrinsic property is of immense value in building efficient recommender systems for e-commerce applications that present consumers with recommendations they are likely to have a higher acceptance. The results of experiments based on this model compare favorably with those from the standard user-based algorithm.