STPR: Personalized Recommendation via Matrix Factorization with Social Ties

In the present situation, the massive rise of Internet platforms and the proliferation of various social network-based websites leads to information overload, users face difficulties in making choices as well as consuming users' time to filter out the products they want. In this context, recommender systems have been proposed to successfully address the information overload. It has been even used to solve the user's cold start problem by recommending products that have been tasted by the user's trusted people under the user's social network ties. In our paper, a personalized recommender system based on social ties-aware to solve the problems of data sparsity, user cold-start, and information overload is proposed. Experiments show that this model effectively reduces computational complexity and improves prediction accuracy compared to the model of the previous works of literature.