A Novel Multi-objective Network Recommender Algorithm Based on PSO

ABSTRACT

This paper models the process of a recommender system as a multi-objective optimization problem, a discrete particle swarm optimization framework is established and has been integrated into multi-objective optimization, consequently, and a multi-objective 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.

Keywords: