Web Service Selection based on Fuzzification and Particle Swarm Optimization

Abstract

In current scenario the information technology acts as a business enabler for industries and organizations. They use services to realize or implement their business process requirements. The service should be selected from collection of compliant services using service discovery based on functional and non-functional requirements. The contemporary service ranking methodologies rank services using statistical or fuzzy methods to select relatively better services. The existing Service discovery is extended to improve the service selection. The proposed system combines Fuzzy, Normalization and Particle Swarm optimization approach to provide the best service as per user requirement. The system initially extracts system Quality QoS attribute values of all complaint service and constructs a matrix. This matrix is transformed into a fuzzy judgment matrix using fuzzy limits. The scores of service are calculated using normalized fuzzy values and Particle Swarm Optimization technique. The experiments are conducted using a set of benchmark web services and concludes that the proposed web service discovery process is more accurate and ranks relatively better service than the existing service ranking methodologies. The accuracies of all hybrid approach are measured using F-Measure and Normalized Discounted Cumulative Gain methods