Goal-based constraint driven dynamic RESTful web service composition using AI techniques

Abstract

SOAP and REST are two architectural styles to develop web service where SOAP follows operation centric and REST follows resource centric approach. Objective of dynamic web service modeling is to compose existing web services published on internet and generate new value-added service to satisfy client request; and it receives great response from many researchers. Even though widespread adoption of web service, there are some situation where we need automatic composition of existing web services from service repository and generate workflow automatically to fulfill client objective. After studying many research papers we find that Automatic RESTful Web Service Composition is still unexplored as compare to SOAP/WSDL-based web service. In this research paper, we proposed goal based constraint-driven internal web service composition and automatic workflow (GCDIWBC) algorithm Using AI planning which generates new value-added service by considering nested automatic service composition and generate workflow also. We also considered case where one automatic web service composition can call other automatic web service composition internally in the composition and generate workflow dynamically which will be run on workflow execution engine like WS-BPEL or WSMO. In other words service dynamically generated by GCDIWBC can nested to other generated service by other. GCDIWBC has implemented using prototype of travel scheduling scenario and performance evaluation with RESTful and SOAP/WSDL web services using state-of-the-art JSHON2 AI planner. Result shows that proposed algorithm with RESTful web service using JSHON2 outperforms SOAP/WSDL standard both in its response time and scalability.