Service technology, driven by its SOA architecture and Web Services, is rapidly maturing, with organizations worldwide adopting this paradigm to publish, request, and compose their applications as web services. However, to remain competitive, these service-oriented applications must be highly composite, adaptive, knowledge-intensive, and reliable. Current Web Services standards like WSDL, BPEL, and WS-CDL provide only static, manual, and process-centric techniques, which are inadequate for deploying such complex services. Moreover, existing research aimed at enhancing these standards for greater correctness and adaptability is still in its early stages and lacks scalability for widespread use. Applications in sectors like E-commerce, E-Banking, and E-health require high adaptability and dependability, often governed by volatile rule-centric knowledge. This thesis aims to improve the development of service-oriented applications by focusing on reliability, dynamic adaptability, and knowledge intensity. Through a comprehensive study and critical analysis of the current state of the art, it proposes an innovative, stepwise approach for engineering and deploying dynamically adaptive, rule-centric service-oriented applications, beginning with the intuitive elicitation of structural service features using stereotyped service-based UML class diagrams.
