A Novel Personalized Recommendation Model Based on User's Context

With the rapid development of information and the Internet technology, people are gradually entering the age of information-overload. Therefore, how to acquire valuable information from the complicated data has become an urgent challenge. Recently, Different personalized recommendation models and algorithms have been proposed to resolve this problem. However, traditional recommendation methods only focus on how to associate items with user interests in a more effective way, while ignoring the user's contextual information. Therefore, it is imperative to coalesce the user's context with the recommendation algorithm together and promote the performance of the recommendation system. In this study, a new personalized recommendation model is proposed to dig latent preferences of users toward context and items, which efficiently integrates the context-aware information into the system. Then a latent context preference recommendation method (LCP-RM) is designed. At last a Standard Gradient Descent method is used to optimize the recommendation model. The simulation results show that our proposed method has achieved the optimal performance in multiple evaluation indicators compared with other algorithms.