User-Based Collaborative Filtering Multi-criteria Recommender System Based on Interaction Between Criteria, Criteria Set with Choquet Integral

The exploitation of knowledge in stored data is one of the current research trends. The increasing of the need of users about searching information, the consulting system is special attented by researchers. Many decision-making solutions for multi-criteria recommender model have been proposed. However, with the intrinsic of the data, the latent values of the interaction relationship, the dominance between the criteria always changes the results of decision-making to advise users. When we take enought these values, decision-making becomes more efficient. In this paper, we propose a new approach to building a decision model for a user-based multi-criteria filtering system with Choquet integration. The operation is also based on the capacity function of a criterion, a set of criteria. The effect of decision making is the degree of interaction between the criteria in the data. This model is also based on traditional techniques and integrates some our new methods. We tested and evaluated the proposed model on the multirecsys tool we built. The standard datasets is used to test. We compare results with some the same existing models. Through experimentation, we saw that the proposed model is quite effective and reliable. It can be applied well in many appropriate systems, contributing to improving the deficiencies and the limitations of the current recommendation.