Approaches for Discovery, Selection and Composition for Heterogeneous Web Services Using Semantic Web

Abstract

A significant amount of research has been progressing to discover, select and compose het- erogenous Web services, including, SOAP, RESTful and Cloud services to fulfill the user request due to the availability and wide acceptance of the Web services. Use of Semantic Web plays a key role to make the service discovery, selection and composition, dynamic and automated by reducing the intervention of the end user. Traditional SOAP services in- creases XML processing overhead at the description and communication level. This issue could be resolved by RESTful services, which are based on the REST (Representational State Transfer) architecture, which makes it more suitable compared to the SOAP services. Meanwhile, cloud services has gained the importance because the cloud computing pro- vides an efficient, on-demand and scalable platform to integrate the resources over the Web from the multiple sources. Very few efforts have been made to develop solutions to perform the service discovery, selection and composition processes in an integrated manner using the Semantic Web. An integrated approach for service discovery, selection and composition using the domain ontology and QoS support for SOAP services has been proposed by me. I have developed a prototype for the Healthcare Information System (HIS) using the proposed approach of SOAP service. For the RESTful services, to perform the service discovery, selection and composition, Linked Open Data-based approach has been proposed by me by considering the QoS parameters (i.e., throughput and response time) and semantic description (i.e. RDF data). Prototypes for the Population Information System(PIS) and Healthcare Recommendation System(HRS) have been developed to demonstrate the effectiveness of the proposed work. I have proposed framework and approaches to perform the cloud service discovery, selection and composition using cloud ontology. Based on the proposed framework the prototype for Healthcare Decision Making System (HDMS) is developed by me. The experimental results show that the proposed approaches have demonstrated effective performance improvement in comparison with existing QoS-based and Semantic Web-based approaches. The proposed solution, fulfills the need of an end user to search, select and compose heterogeneous Web services with increased level of user satisfaction and higher degree of automation.