Meta-recommender approach using meta-knowledge

Generally, the user preferences change over time which implies the need to adapt the recommendation technique used in order to recommend pertinent items. For this reason, in this paper we propose a meta-recommender that follows the change of user's interests over time to propose the appropriate recommendation technique. The proposed approach is based on meta-knowledge called explanation and a hybrid approach, and has two main phases: the first phase is to fill the meta-knowledge database with explanations using a hybrid recommendation approach. In the second phase, we calculate the average of each explanation for a user to determine the recommendation technique to use.