A Combinatorial Model Based on Users’ Interest in Order to Improve the Performance of the Recommender Systems

Nowadays there is an increasing interest in the recommender systems, which recommend various items such as music, film, book, etc. Applying these systems which needs customers recognition, we lead to some suitable and high quality services in order to provide and sell some products aiming at the increase of customer satisfaction, the increase of the rate of future returns, the increase of customer loyalty, acquiring highest possible level of sale and to gain the credit and validity. Two steps are presented here to provide recommender system; the first step is providing ranked items to users, based on a user's choice of items and identifying common interests of him and other users. In order to improve the performance of the system, we provide a combinatorial method to combine methods of the first step.