Large-Scale Web Service Collection for Service Archive Management

Abstract

The adoption and proliferation of Web services lead to an increasing number of Web services emerging on the Internet. Therefore quickly and accurately discovering services that satisfy the needs of service consumers becomes difficult. There is an urgent need to manage Web services effectively. From a holistic and historical perspective, this paper proposes a Web service archive which describes and records the lifecycle of Web services that ever published on the Internet, rather than merely stores service information that are currently available as the conventional service registries do. To build up the Web service archive, our framework collects and integrates a large scale of services scattering all over the Internet, and enriches multiple aspects of service information from WSDL documents and related Web pages. On the basis of the service archive, we analyze the spatial and temporal distribution of archival metrics, QoS measurements, and history records of Web services. Our approach can periodically keep track of Web service changes and service lifecycle. In addition, experiments are conducted using 15372 crawled services in the real world. Our approach provides insights on investigating the state of the art in Web service technologies, finding technology trends, and providing valuable data for future research.