At present, the demand for consultation for users is increasing with diverse information. Multi-criteria recommender system is one of the research goals of scientists that is of great interest. Many recommender methods are designed to find the most valuable products or services that suggest to user best consultancy. Selecting a suitable solution for recommendation on data storage will response well the requirements of the users. In this paper, we propose a new approach to build multi-criteria recommender model that interacts based on items-based collaborative filtering using the ordered weighted average operator on sparse datasets. This model demonstrates the coherence and impact of user criteria in decision-making. The model was evaluated empirically on the multirecsys tool on three datasets: MovieLens 100K, MSWeb and Jester5k. The experiment also illustrates the comparison with some other researched methods that applied. Consultancy results of the proposed model are quite effective compared to some traditional consulting models. This counseling model can be applied well in a variety of contexts. Especially, in the case of sparse data, the counseling results of the proposed model seem always better than the exiting models (item based).