| A major advantage of the emerging service-oriented computing technology is its ability to enable rapid and flexible development of large-scale distributed systems achieving various system goals. Such service-based systems are composed of loosely coupled services implemented in various programming languages, deployed over multiple service platforms across various organizational boundaries and over heterogeneous networks. Service matching is a basic critical building block in service-based systems. It mediates between service providers and service requesters to efficiently locate suitable services to perform dynamic tasks. Little work has been done in traditional syntax-based and parameter-based service matching approaches to handle the rich semantics of services. Advanced semantic-based service matching capabilities are desirable to provide more precise and relevant matching results.; In this dissertation, a semantic-based service matching approach is presented to provide accurate, relevant, privacy-preserving and efficient service matching capabilities. This approach is based on the semantic modeling, specification, interpretation and analysis of service functionalities, attributes, situational data and access control policies. This approach first extends existing service semantic model OWL-S with situation, access control policy, and service discovery ontologies to model and specify more complete service semantics for service matching. By interpreting and analyzing the service semantics specified based on the modeled ontologies, this approach ensures that the matching results will have compatible functionalities, similar characteristics and relevant situational data to the service discovery request. A privacy-preserving mechanism is also incorporated in this approach to protect the shared privacy-related information from being leaked to unauthorized parties during service matching process. |
