Files
Abstract
Since the introduction of Web services and the service-oriented architecture, they have been widely adopted in industry and academic settings. Web services have the advantages of standardization of service interfaces and facilitating system integration and cross-platform interoperability, which are extensively utilized by business process systems and scientific workflow systems. In the research area of model-driven software engineering, even the simplest software system needs to address the major issue of requirements and constraints during the whole lifecycle. Web service-based business processes and scientific workflows are no exception. Process constraint modeling and verification, focusing on how to enforce the conformity of process constraints throughout its lifecycle, including design, deployment and runtime execution, remains a big challenge, especially when such constraints are considered in the composite Web services and workflow systems. In this dissertation, we propose a semantic end-to-end Process Constraint Framework (PCF) to address this issue. The framework includes a process constraint model, consisting of a process constraint ontology (ProContO) and a process constraint language (PCL). In addition to constraint modeling at design time covered by the representation model, both static and dynamic verification in process constraints lifecycle are addressed in PCF. While the static verification of process constraints concentrates on syntactic, semantic and service specification verifications during design and deployment phases, dynamic verification focuses on the runtime process execution with the help of the underlying monitoring module. Moreover, PCF supports both the definition and implementation of user-defined exception handling methods, as exceptions are ubiquitous in the distributed environment. Three motivating scenarios: (i) prevention of nepotism in job interviews, (ii) safe distance guarantee in tornado emergency reaction process, and (iii) selection of service network location in a glycomics scientific workflow process, are used to illustrate how ontology and semantic technologies are utilized to model different kinds of process constraints in their whole lifecycle.