Personalized Paper Recommendation Based on User Historical Behavior
Communications in Computer and Information Science(2012)
摘要
With the increasing of the amount of the scientific papers, it is very important and difficult for paper-sharing platforms to recommend related papers accurately for users. This paper tackles the problem by proposing a method that models user historical behavior. Through collecting the operations on scientific papers of online users and carrying on the detailed analysis, we build preference model for each user. The personalized recommendation model is constructed based on content-based filtering model and statistical language model.. Experimental results show that users' historical behavior plays an important role in user preference modeling and the proposed method improves the final predication performance in the field of technical papers recommendation.
更多查看译文
关键词
Personalized Recommendation,User Historical behavior,Similarity,Recommendation Model
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络