A Personalized Recommendation Algorithm for User-Preference Similarity Through the Semantic Analysis

作者：Zhang Haolin^*; Ye Feiyue

来源：17th International Conference on Web Information Systems Engineering (WISE), 2016-11-08 to 2016-11-10.

DOI：10.1007/978-3-319-48740-3_22

• Personalized recommendation
• Semantic analysis
• User-Preference
• Project-Relevancy

Traditional personalized recommendation algorithms do not involve the analysis of semantic information, so the recommendation results are less accurate. Aiming at this problem, based on the semantic analysis, genres similarity and content features similarity of the projects rated by users are used to measure user-preference and therefore calculate users similarity. Moreover, the number of projects in the same genres is applied to measure project-relevancy and thereby project similarity. Based on these, this study puts forward a personalized recommendation algorithm for user-preference similarity through the semantic analysis. The contrast experiment results based on Movielens data set show that the recommendation accuracy and quality of the proposed algorithm are significantly improved.

• 出版日期2016
• 单位上海大学