A Recommendation Mechanism Combined with Bayesian Networks and Incentive Theory - A Recommend System Design

In recent years, due to the development of e-commerce, recommendation systems have been widely used in many fields. Although the recommendation technology has become increasingly mature, there are still some problems. For example, the recommendation system cannot provide recommendation information correctly under the premise of insufficient user information. This paper proposes a recommendation mechanism combining Bayesian network and incentive theory, by finding out the potential rules in the user profile and encouraging users to share the evaluation of new items under the premise of individual rationality and incentive compatibility. Experiments have shown that expanding the user profile and encouraging users to rate can help the system improve the recommendation accuracy of the system in the case of insufficient user information.