An enhanced personalized recommendation utilizing expert's opinion via collaborative filtering and clustering techniques
M. Sridevi, R. Rajeswara Rao · 2016
The emergence of web had a profound impact on personalized Recommender systems. They empower most of the popular e-commerce websites by changing the way the users communicate with the websites. Despite the available of popular recommendation techniques like Content based and Collaborative filtering which have been successful in implementing satisfactory systems, they still have some downsides. One of the most notable issue is to generate efficient recommendations when there is sparsity of data. Aimed to address the issue to some extent, this paper proposes a framework that integrates the collaborative filtering features and clustering techniques. Items and Users are clustered separately using similarity techniques from which a user is promoted as an expert from each user group based on the user whose profile is more concentrated on a particular class of items. The proposed method aims in delivering accurate recommendations by considering the experts opinion when there is sparsity of data.