基于服务相似度的API服务推荐方法研究  

Study on the recommended method of API service based on service similarity

作　　者：李泽平[1] 

机构地区：[1]黔南民族师范学院计算机科学系,贵州都匀558000

出　　处：《贵州师范学院学报》2013年第6期28-32,共5页Journal of Guizhou Education University

基　　金：贵州省教育厅自然科学类科研项目资助(项目批准号:黔教科20090048);黔南民族师范学院院级教改项目资助(项目批准号:jg-11-02)

摘　　要：利用API服务和Mashup应用的注册和组合信息,通过构建Mashup-API网络和API-API网络,在LFH方法的基础上改进Rada&Courtney算法,进行API服务的标签相似度、描述相似度和摘要相似度的计算并最终计算其服务相似度,基于服务的使用场景提出了相应的服务推荐算法,实现了协作关系、竞争关系和其他关系的服务推荐。使用ProgrammableWeb上Mashup应用和API服务的真实数据为载体验证了方法的可行性和有效性,在服务分类和服务推荐方面具有一定的意义。Using the API Service and Mashup applications for the registration and portfolio information, through constructing Mashup - API and API - API network, and on the basis of the LFH method, the paper improves Rada ＆ Courtney algorithm and calculate the tag similarity of API services, descriptive similarity, abstract similarity, and finally its service similarity. Based on the service usage scenarios, the paper provides corresponding service recommendation algorithm, and realizes service recommendation for cooperative relationship, competitive and other relationships. Mashup application and the real data of API service from ProgrammableWeb are used to verify the feasibility and effectiveness of the method for the carrier, which has certain significance in the classification of services and the services recommendation.

关 键 词：API服务 服务关系 服务相似度 服务推荐 验证 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]