An interactive evolutionary approach to designing novel recommender systems

Recently, recommender systems have been widely used in e-commerce websites to help customers discover items they want. Since a recommender system should be able to provide users with helpful information regarding items that might interest them, the ability to immediately respond to changes in a user’s preference is a valuable asset of the systems. Thus, this work presents a novel recommender system that effectively adapts and immediately responds to any changes in the system by utilizing content-based filtering within the framework of interactive evolutionary computation. In addition, a data grouping technique is employed to enhance the computational time efficiency. The proposed system is then tested with music data. Experimental results confirm that the proposed system is able to offer users suitable music items with assured quality in a timely manner. Furthermore, the experimental results suggest that this system provides more reliable music recommendations than other contentbased filtering systems even in the working environment changes.