基于协同过滤的可信Web服务推荐方法

发布时间：2018-02-17 00:18

本文关键词： Web服务协同过滤非功能需求可信服务相似用户　出处：《计算机应用》2014年01期 　论文类型：期刊论文

【摘要】：为了实现对可信Web服务的推荐,在分析了Web服务推荐技术与电子商务推荐技术的不同的基础上,提出了一种基于协同过滤的可信Web服务推荐方法。首先,根据已有成果对待组装应用的可信需求进行评估,根据此需求对相似用户进行首次筛选;然后在首次筛选的用户中,根据用户使用服务后的评分数据和用户信息来对相似用户进行二次筛选,经过两次筛选得到最终推荐用户。在根据用户对服务的评分数据计算用户之间的相似性时,考虑了不同服务对于用户间相似性的贡献值;在根据用户信息计算用户之间的相似性时,考虑到用户信息之间非线性的特点,引入了欧几里得距离公式来计算其相似值;在产生推荐的过程中还考虑了不诚实用户和用户数不足的问题。模拟实验结果表明该方法能够有效地对可信Web服务进行推荐。
[Abstract]:In order to realize the recommendation of trusted Web services, based on the analysis of the difference between Web services recommendation technology and e-commerce recommendation technology, a trusted Web service recommendation method based on collaborative filtering is proposed. Evaluate trusted requirements for assembly applications based on existing results, and screen similar users for the first time based on this requirement; then, among first-time filtered users, According to the users' rating data and user information after using the service, the similar users are screened twice, and the final recommended users are obtained after the second filtering. When calculating the similarity between users based on the rating data of the service, The contribution of different services to the similarity between users is considered, and the Euclidean distance formula is introduced to calculate the similarity between users, taking into account the nonlinear characteristics of user information. The problem of dishonest users and insufficient number of users is also considered in the process of producing recommendation. Simulation results show that the proposed method can effectively recommend trusted Web services.
【作者单位】：云南大学软件学院;云南省软件工程重点实验室;云南大学信息学院;云南大学经济学院;
【基金】：国家自然科学基金资助项目(61379032,61262024,61262025) 云南省应用基础研究计划项目(2012FB118,2012FB119) 云南省教育厅科学研究基金资助项目(2012Y257,2012C108) 云南省软件工程重点实验室开放基金资助项目(2011SE09) 云南大学“中青年骨干教师培养计划”专项经费资助项目
【分类号】：TP393.09

【参考文献】