基于企业用户需求的科技服务资源综合推荐算法研究

Enterprise User Requirement-based Scientific and Technological Service Comprehensive Resources Recommendation

摘要:

[研究目的]现有科技服务平台中科技服务资源数量指数级增长、服务质量多样化,以及企业用户需求难以量化。为解决科技服务平台企业用户与服务资源间的精准匹配问题,提出一种基于企业用户需求的科技服务资源综合推荐算法(EURSTS)。[研究方法]综合考虑科技服务特殊属性和企业背景信息,采用模糊模型量化信息和综合相似度求解,对企业用户和科技服务资源进行匹配推荐。[研究结论]通过与CB算法、基于服务QoS算法的对比说明EURSTS算法能显著的改善推荐效果,其准确率平均提升了30.1%~37.1%,召回率平均提升了0.1%~7.9%,验证了该算法的有效性。

Abstract:

[Research Purpose] In the existing science and technology service platform, the number of science and technology service resources has increased exponentially, the service quality is diversified, and the needs of enterprise users are difficult to quantify. In order to solve the problem of accurate matching between enterprise users and service resources of science and technology service platform, a comprehensive recommendation algorithm of science and technology service resources (EURSTS) based on the needs of enterprise users is proposed. [Research Method] Considering the special attributes of science and technology services and enterprise background information, the fuzzy model is used to quantify the information and solve the comprehensive similarity, so as to match and recommend enterprise users and science and technology service resources. [Research conclusion] Compared with CB algorithm and service QoS algorithm, it shows that the EURSTS algorithm can improve the effect of recommendation significantly, the average accuracy rate is improved by 30.1% ~ 37.1% , the average recall rate is improved by 0.1% ~ 7.9%, the performance of the algorithm is verified.