2020/09/10 by Saman Forouzandeh, Forouzandeh, Saman, Mehrdad Rostami +3
Computer Science · #FOS: Computer and information sciences #Machine Learning (cs.LG) #Recommender Systems and Techniques #Sentiment Analysis and Opinion Mining #Social and Information Networks (cs.SI) #Spam and Phishing Detection
paper · pdf · doi:10.48550/arxiv.2009.04825
openalex publication_date 2020/09/10 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
The use of recommender systems has increased dramatically to assist online\nsocial network users in the decision-making process and selecting appropriate\nitems. On the other hand, due to many different items, users cannot score a\nwide range of them, and usually, there is a scattering problem for the matrix\ncreated for users. To solve the problem, the trust-based recommender systems\nare applied to predict the score of the desired item for the user. Various\ncriteria have been considered to define trust, and the degree of trust between\nusers is usually calculated based on these criteria. In this regard, it is\nimpossible to obtain the degree of trust for all users because of the large\nnumber of them in social networks. Also, for this problem, researchers use\ndifferent modes of the Random Walk algorithm to randomly visit some users,\nstudy their behavior, and gain the degree of trust between them. In the present\nstudy, a trust-based recommender system is presented that predicts the score of\nitems that the target user has not rated, and if the item is not found, it\noffers the user the items dependent on that item that are also part of the\nuser's interests. In a trusted network, by weighting the edges between the\nnodes, the degree of trust is determined, and a TrustWalker is developed, which\nuses the Biased Random Walk (BRW) algorithm to move between the nodes. The\nweight of the edges is effective in the selection of random steps. The\nimplementation and evaluation of the present research method have been carried\nout on three datasets named Epinions, Flixster, and FilmTrust; the results\nreveal the high efficiency of the proposed method.\n