OntoDisco: Improving Web Service Discovery by Hybridization of Ontology Focused Concept Clustering and Interface Semantics

Abstract

Web services can be regarded as the most vital constituent of the World Wide Web as each of the web services is associated with an important role. With the paradigm shift toward Service Oriented Architecture, Service computing has become quite predominant across the globe. Web service discovery is not just important but also crucial and cumbersome. This paper proposes a new web service discovery approach that is semantics aware. This approach helps to extract the underlying semantics and enables users to utilize web services. The strategy focuses on the creation of a web database comprising of dataset crawled from well-known registry sites. A word-level semantics is applied using clustering based on the NPMI value. The library creation happens based on the semantic similarity values. An ontology for the principal classes of web services is modeled to facilitate clustering based on the initial aggregation of the data and the concepts. Finally, a web service discovery approach is designed that furnished related web services based on user queries. The proposed strategy yields an accuracy of 90.4 % which is much better than the existing approaches.