Analysis of Dynamic Web Services: Towards Efficient Discovery in Cloud

Abstract

Web services are very dynamic as they can be added, modified and deleted from the repository on the fly. Discovering web services for service composition have special requirements. In the service composition, service requirement is not only to satisfy its functional specification but it also needs to ensure that the selected composing services are compatible with respect to their input/output structure and its associated data types. Web Service community [1] organized the functionally similar services together in a group. But one important drawback of using web service community in the service composition is that it requires further analysis of the discovered service communities for selecting appropriate service which can be composed based on its input/output structure and data types. This paper proposes data structures, which can be used to create web service family [2] by filtering not only on the basis of functionality but also on the basis of its input/output structure, which are required for composing services with one another. In order to speed up the web service family creation process, the creation of web service family is divided in two phases. The first phase of the proposed approach introduces a novel concept of business family and dynamic business web service (DBWS) tree, which are used to organize all web services in the tree structure on the basis of their business context and input/output structure. In the second phase, the concept of DBWS tree is used further to create the web service families. The DBWS resembles the classification and organization of businesses in NAICS. The DBWS tree can be used very efficiently for creating web service families, which in turn can be used for selecting service efficiently in service composition.