Research On Collaborative Filtering Recommendation Algorithm Based On Trust

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 Collaborative Filtering algorithms focus on recommending the most relevantitems to the users and don’t take the social contextual information into consideration. Theresults may lead to deviating the user’s needs. Therefore, how to improve the quality of therecommendation has become a widely hot issue in the recommendation field. In this paper,we further study the Collaborative Filtering algorithm that incorporates the contextinformation.First of all, aiming at the problem that the traditional Collaborative Filteringalgorithm has low recommendation accuracy, in this part, we propose a collaborativefiltering recommendation algorithm based on the global trust degree integrating the directtrust information in the social networks. We first transform the local trust relationships tothe global trust relationship by the rules in the trust network, and get the trust rank of allusers in the trust networks; Then we use the global trust value to instead of the similarityinformation value as the weights of a predicted formula in the traditional collaborativerecommendation algorithm, and integrate the weights to the matrix factorization-basedrecommendation model.Secondly, aiming at the problems that the existing model-based collaborative filteringalgorithm has low recommendation accuracy and small recommendation coverage, in thispart we propose a collaborative filtering recommendation algorithm based on the trustpropagation by introducing the trust information of social network to extend the matrixfactorization-based recommendation model. We first design a set of trust propagation rulesbased on the direct trust relationships of the social network, so as to propagate the trustrelationship in the social networks, and get to quantize the new trust relationship. Then weload the quantitative trust relations after the trust propagation as the trust weight into the matrix factorization-based model according to the characteristics that the matrixfactorization technique can reduce the dimension of large-scale datasets.Finally, we give the experimental evaluations and analysis of the algorithmsproposed in this paper, compare the performance between the proposed algorithms andother existing algorithms.