基于Vague值的电子商务推荐系统及其相似度研究

Recommender Systems of EC and Its Similarity Based on Vague Value

针对目前电子商务推荐系统中存在的核心问题———相似度,提出借助Vague集理论研究推荐系统的思想。电子商务过程中顾客行为不确定性的存在,为Vague集的引入提供理论基础。商品推荐依赖的是商品间或顾客间的相似程度,而相似度的计算正是Vague集研究较为成熟的一个领域。根据一般电子商务购物方式,确定不同的顾客类型,在顾客分类的基础上,利用统计方法定义商品的Vague值,实现电子商务推荐系统与Vague的完美结合,并通过相似度的计算验证该方法的有效性,从而为推荐系统的研究提供新的思路和方法。

In the paper,vague set theory is introduced into the study of recommender systems to solve its core problem which is similarity.The existence of uncertainty of customer behavior in the course of e-commerce provides a theoretical basis for the introduction of Vague set.Recommendation of goods relies on the degree of similarity between customers or goods,while the calculation of similarity is a mature area in the research of Vague set.First,Different customer types are identified according to the general shopping way in e-commerce.Then based on the customer classification,statistical methods are used to define the Vague value of the commodity.This method makes a perfect combination of e-commerce recommendation system and Vague set,and provides new idea for the study of e-commerce recommendation system.