A Multi-interests Model of Recommendation System Based on Customer Life Cycle

Recommendation System is one of the most important research topics in E-commerce, and multiple recommendation methods have been developed. However, most traditional ways didn't consider the variability and diversity of user interests. To solve this problem, this paper introduces the Customer Life Cycle theory of Customer Relation Management to analyze changes of user interests. Further, we define 2 measurement indexes, Interest Depth and Interest Span, and propose a multi-interests model of Recommendation System based on Customer Life Cycle. Experimental results show that our model is able to track changes of user multiinterests and help recommendations.