Collaborative Filtering Recommendation Algorithm Incorporating Context Information

At present, as an essential method of information filtering technique, collaborativefiltering recommender systems have attracted an extensive attention. With thepopularization of the Internet, the scale of the electronic commerce system become moreand more big, and the number of the user and item begin to increase sharply. However, thetraditional CF algorithms focus on recommending the most relevant items to the users anddon’t take the contextual information into consideration. The results may lead to deviatethe user’s needs. Therefore, how to improve the quality of the recommendation havebecome a widely hot issue in the recommendation field. In this paper, we further study theCF algorithm that incorporates the context information.First of all, aiming at the problem that the change of user preferences can’t bereflected in time, we propose a recommendation algorithm based on user ratings time. Weassign the information of user rating time in the process of the recommendation, andpropose a context-aware recommendation algorithm incorporating time information.Firstly, a calculating method of the time weight is presented. According to the user’s ratingtime, we assign a different time weight to each rating, in order to reflect the dynamicchanges of the user’s interests. Then the time weight is incorporated into the basic matrixmodel and the data sparsity is solved effectively. The algorithm can significantly improvethe accuracy of the recommendation results.Secondly, aiming at the traditional CF algorithms ignore the contextual informationof the users, a context-aware recommendation algorithm based on Fuzzy kernel clusteringis proposed. Firstly, the rating contextual information are clustered by Fuzzy kernelclustering algorithm, for generating the clusters and the membership matrix; then thetarget user’s current context match with the context cluster center, and the rating data ofnon-active context is mapped to active context clustering; finally choosing the user’srating information which meet the target user’s current context information to make therecommendation.Finally, the effectiveness of the algorithm is verified by experiment, and theperformance between the proposed algorithm and the existing algorithms are compared.