引入漂移特性的用户兴趣模型优化研究

Research on User Interest Model Optimization Introducing Drift Characteristics

根据用户浏览网页时的操作行为,通过量化的方法建立起用户兴趣模型来反映用户兴趣,从而针对不同用户推荐其可能感兴趣的文章。基于兴趣模型的更新效率问题和用户兴趣的漂移特性,引入兴趣模型的时间分段机制和时间衰减机制,对兴趣模型进行了持续优化。实验表明,优化的兴趣模型在系统性能上有较大的提升,并能较好地反映出用户的兴趣变化,对于用户兴趣的表征更加准确,从而进一步提高了兴趣模型推荐文章的准确率。

Users' interest model is set up to reflect users' interest with quantitative methods when users are browsing webpage. The interest model is used to recommend articles to users later. On the basis of the preview studies, user interest model in this paper introduces time slicing and time decay mechanism to improve update efficiency and reflect interest drifting. Experimental results show that the optimization of user interest model obviously improves the system efficiency and better reflects users' interest drifting, which further improves the accuracy of the interest model recommended articles.