一种基于贝叶斯网络的个性化协同过滤推荐方法研究

Method of Personalized Collaboration Filter Recommendation Based on Bayesian Network

针对推荐系统不能有效进行个性化推荐问题,在协同过滤过程中引入语义校验,通过对基于用户的协同过滤推荐结果进行语义校验,剔除概率较低的推荐结果,选择概率较高的结果推荐给用户,从而实现个性化语义推荐。在构建贝叶斯语义校验网络时,增加用户"喜好"偏好字段,通过问卷调查及信息反馈,确定用户对物品的喜好偏好值,确保贝叶斯语义校验网络的科学性。实验结果表明,本方法能剔除用户喜好度较低的物品,提高用户的满意度。

Lacking of high efficiency personal recommendation in collaboration filtering recommendation system by using semantic recommendation system, we proposed a new method in checking. We check the result of collaboration filtering based on user item by Bayesian semantic to eliminate the item of lower probability, and to select the higher probability item to users. In constructing the Bayesian semantic check network, we add an emotion field named ＂fancy＂ by question- naire survey and information feedback, and we decide user＇s emotion for some goods to ensure the scientific of semantic checking network. Experiments show that the method can eliminate the items with low user preferences and improve the satisfaction degree of the users.