工业工程 ›› 2019, Vol. 22 ›› Issue (2): 1-9.doi: 10.3969/j.issn.1007-7375.2019.02.001

• 专题论述 •    下一篇

面向用户个性化需求的云服务商推荐

禹春霞, 刘玉茹, 高旭   

  1. 中国石油大学(北京) 经济管理学院, 北京 102249
  • 收稿日期:2018-10-19 出版日期:2019-04-30 发布日期:2019-04-22
  • 作者简介:禹春霞(1983-),女,山东省人,副研究员,博士,博士研究生导师,主要研究方向为供应链管理、个性化推荐
  • 基金资助:
    国家自然科学基金青年基金资助项目(71501187);教育部人文社会科学青年基金资助项目(14YJC630179)

Cloud Service Supplier Recommendation based on User Personalized Preference

YU Chunxia, LIU Yuru, GAO Xu   

  1. School of Economics and Management, China University of Petroleum-Beijing, Beijing 102249, China
  • Received:2018-10-19 Online:2019-04-30 Published:2019-04-22

摘要: 针对当前不具备专业知识的用户难以从海量云服务中选择满足其偏好的云服务商的问题,构建了满足用户需求偏好的云服务商推荐模型。该模型包括以下3部分:首先,从用户角度,通过模糊评价的方法确定并衡量用户对云服务的需求偏好;其次,从云服务商角度,通过模糊评价法和熵权法确定并衡量其满足用户需求的能力;最后,利用相似距离公式,将用户与候选服务商的相似性程度进行排序,向用户推荐最匹配的云服务商。算例结果表明,与传统的推荐方法相比,该模型能够更好地针对用户对云服务各项指标的偏好进行推荐,提高了用户选择云服务商的准确性。

关键词: 云服务, 云服务商, 需求偏好, 个性化推荐

Abstract: At present, it is difficult for users without professional knowledge to choose a cloud service provider that meets their preference. To solve this problem, a cloud service provider recommendation model based on users' preferences is proposed. The model is composed of three parts. The first part is to determine and measure the user's preferences for cloud services from the perspective of user through the fuzzy evaluation. The second one is to identify and measure suppliers' capacity to meet the user's needs from the perspective of cloud service providers through the fuzzy evaluation and entropy weight method. The third is to measure the similarity of user and candidate suppliers using the similarity distance formula, and provide recommendation to the user. Experimental results show that, compared with the traditional recommendation, the proposed model can recommend cloud service providers that meet the needs of users according to their evaluation of various cloud service indicators, improving the accuracy in the process of choosing cloud service providers.

Key words: cloud service, cloud service provider, user preferences, personalized recommendation

中图分类号: