Domain Specific Web Service Composition by Parameter Classification Using Naïve Bayes Algorithm
T. F. Michael Raj, Kandala N. V. P. S. Rajesh · 2014
2 Abstract: Web service composition for the specific implementation or business process is typically based on the service description and its parameter constructs. Discovering and compositing the WS from the pool of web services is a challenging task and more attention has to be given on the similarities checking. In this paper we propose an approach on identifying the most appropriate service based on the user's preferences of the requested WS. The given WS description may contain the parameters, which may have relations with the requested WS of the specific domain in different aspects like name, parameters and types. The domain specific WS classification can be done using Naive Bayes classification algorithm and we have given the experimental evaluation on the sample WSDL datasets and the simulated results shows that effectiveness and efficiency of the proposed approach.