Context-sensitive Web service discovery over the bipartite graph model

• Published in: Frontiers of Computer Science Vol. 7; no. 6; pp. 875 - 893
• Main Authors: ZHANG, Rong, ZETTSU, Koji, KIDAWARA, Yutaka, KIYOKI, Yasushi, ZHOU, Aoying
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Higher Education Press 01.12.2013; Springer Science and Business Media LLC; Springer Berlin Heidelberg; Springer Nature B.V
• Subjects: bipartite graph model; Computer Science; Context; context-usefulness; Electronic documents; Graph theory; Iterative algorithms; Iterative methods; Queries; Query expansion; Query processing; Research Article; Search engines; Service oriented architecture; usage context; Web service; Web services; Web服务发现; 上下文; 图模型; 敏感; 服务内容; 服务空间; 查询处理; 面向服务的架构; usage context; bipartite graph model; context-usefulness; Web service
• ISSN: 2095-2228; 2095-2236
• DOI: 10.1007/s11704-013-1256-x

As service oriented architecture （SOA） matures, service consumption demand leads to an urgent requirement for service discovery. Unlike Web documents, services are intended to be executed to achieve objectives and/or desired goals of users. This leads to the notion that service discovery should take the ＂usage context＂ of service into account as well as service content （descriptions） which have been well explored. In this paper, we introduce the concept of service context which is used to represent service usage. In query processing, both service content and service context are ex- amined to identify services. We propose to represent ser- vice context by a weighted bipartite graph model. Based on the bipartite graph model, we reduce the gap between query space and service space by query expansion to improve re- call. We also design an iteration algorithm for result ranking by considering service contextsefulness as well as contentrelevance to improve precision. Finally, we develop a service search engine implementing this mechanism, and conduct some experiments to verify our idea.