Reserch On Dynamic Trust Network Based Context Aware Recommendation Algorithm

In the field of recommender systems, trust network based context aware recommendation algorithm has been widely studied, in addition to the algorithm considering the binary relationship between users and items, researchers also have proposed a series of improved methods to consider information trust and contextual information. However, these algorithms take the trust information into account, ignoring the attenuation of trust in the transfer process. Therefore, to deal with the question of dynamic change of trust in the transfer process, we further study the collaborative filtering algorithm in this paper.First of all, according to the problem of implicit trust network construction in the existing trustbased collaborative filtering method, using the technologies of trust network, trust propagation, we propose the user-based implicit trust network model and item-based trust network model from the aspect of user and item respectively.Secondly, in order to solve the problems of the attenuation of transactions in the existing trustbased collaborative filtering method, we use the social network to construct the trust network and find the best paths between users, then we will make the reasonable trust attenuation on the paths.Finally, we conduct comparative experiments with the existing two kinds of recommendation algorithms on a real dataset and three semi-synthetic datasets. Experimental results show that the four algorithms proposed in this paper outperform the existing two kinds of recommendation algorithms in the prediction performance.