Multi-objective Particle Swarm Optimization Algorithm for Recommender System

This paper models the process of a recommender system as a multiobjective optimization problem, a discrete particle swarm optimization framework is established and has been integrated into multiobjective optimization, consequently, a multiobjective discrete particle swarm optimization algorithm is proposed to solve the modeled optimization problem. Each run of the current mainstream recommender algorithms can only give one recommendation scheme, while the proposed algorithm can generate a set of schemes, and it provides the decision maker with more choices, a decision maker can make use of users' registration information to choose a personalized scheme to recommend to the user. In order to check the performance of the proposed algorithm, experiments on simulated and real-world data are tested and a comparison with the classical user based collaborative filtering recommendation method is made, experiments demonstrate the effectiveness of the proposed algorithm.