Clustering-Based Web Services Management

Internet has changed and is changing our world here and there today. In Internet saturated with all kinds web services, applications such as web sites is powerful much more than before to perform various kinds of tasks and provide satisfied services for users. In service oriented Internet environments how to coordinate all kinds of web services to cooperate with each other is facing new challenges different from what traditional workflow technology performs.With the development of Internet, semantic web service comes up with the combination of web service and semantic web. And there are a large number of web services and semantic web services on the web available. It is a challenge for management of web services.Nowadays, UDDI is a common way to manage web services for supporting publishing, finding functionality with the keyword searching and standard classification. It is not fit for the recall and precision of web services with the development of web services.So we propose an clustering-based web service management architecture for supporting the web service management. Based on the ontology vocabulary and web service similarity computing method, the architecture uses the clustering method to group the registered web services according to their functionality, extracts the web services in the same group and encapsulates a "abstract web service" to stand for the web services. The architecture uses the "abstract web service" to support web services register and discovery.Based on the above, the paper proposes CWSMA (Clustering-based Web Service Management Architecture), and applies the architecture in the IPVita platform, which is based on the SmartFlow Pattern. IPVita discovers web services registered in the platform and makes a satisfactory travel plan for the users according to their requirements and the smart flow technology. In here, IPVita uses the CWSMA to give a better performance on web services recall and precision.