A user-oriented content based recommender system based on reclusive methods and interactive genetic algorithm

Abstract

Due to the unprecedented proliferation of textual contents such as blog articles, news, research papers and other things like movies, books and restaurants etc. on the web in recent years, the development of content-based recommender systems (CB-RSs) has become an important research area which aims to provide personalized suggestions about items to users while interacting with the large spaces on the web using items' contents and users' preferences. However, item representation and responds to changing user preferences are still major concerns. In our work, we have employed Reclusive Methods (RMs) to deal with the uncertainty associated with item representation and Interactive Genetic Algorithm (IGA) is used to adapt the system with changing user preferences. First of all, a fuzzy theoretic approach to content based recommender system (FCB-RS) is designed to generate initial population for IGA using reclusive methods. Second, K-means algorithm is employed for clustering the items in order to handle the time complexity of IGA algorithm. Thereafter, a user-oriented content based recommender system (UCB-RS) is developed by incorporating IGA into FCB-RS through user evaluation. Experimental results show that the proposed system (UCB-RS) outperforms both the classical CB-RS and the FCB-RS. © 2013 Springer.