Knowledge Support for Software Processes

Abstract

Last decades of software development discipline have led to creation of sophisticated software engineering methods that support companies and help them to be on par with very competitive software business environment. Supporting methods mentioned above are based on business processes known as software processes or software engineering processes. This thesis presents a combination of visual process modeling combined with formal knowledge representation support and enterprise integration in organizations. The thesis gives an overall insight to existing frameworks, languages, and methods generally used to capture knowledge, and processes and compares them with proposed solution. Author based the proposed software process framework on semi-formal and formal process modeling foundations with terminologically strong process meta-model behind; this combination offers rich modeling capabilities of semi-formal methods extended with formal knowledge support. Proposed solution is concluded with a methodology that meets desiderata of software process and knowledge capture and modeling. The methodology over the formal knowledge framework for process representation includes modeling paradigms and best practices that allow iterative approach, distinction of process models (static and dynamic), and enables simplified models creation, adaptation, refactoring, and verification. Case study showing features of the framework and application of the methodology is based on requirements management processes in a domain of healthcare information systems development.