Understanding and Supporting Knowledge Sharing in Software Engineering

Very often in practice, the problems in software engineering projects seem to have their origins in knowledge sharing difficulties. This creates a need to understand knowledge sharing in the software engineering field better and to find ways to support it. Through an improved understanding of knowledge sharing and especially the problems in it, the targets for support can be determined. By supporting knowledge sharing the maturity of software development processes can be significantly raised. This thesis presents new approaches and offers tools to understand and support knowledge sharing. The main results are the Knowledge Sharing Framework (KSF) as a tool for understanding and the Knowledge Sharing Pattern Language as a tool for supporting knowledge sharing. The purpose of the KSF is to help to identify the current knowledge sharing situation in an organization. This creates the basis for improvement actions in an organization. In this thesis the KSF is utilized in an empirical case study as well as a being framework for understanding differences in knowledge sharing approaches in different software development methods. KSF profiles are used to describe the results of KSF analyses in an illustrative way. After gaining more understanding through defining and using the KSF, the focus is transferred to how to support knowledge sharing. A technique, knowledge sharing patterns, is proposed here to improve existing software development processes with better knowledge sharing properties. Knowledge sharing patterns describe solutions to problems frequently related to the difficulties in knowledge sharing and offer a light-weight mechanism to introduce knowledge sharing sensitive practices in development processes. The knowledge sharing patterns form the Knowledge Sharing Pattern Language that covers several important knowledge sharing practices in a software development organization. The process for creating and evaluating Knowledge Sharing Pattern Language is one of the results. It is a process of creating an organizational pattern language that is applicable also in other contexts. The resulting Knowledge Sharing Pattern Language never will be complete. A pattern language needs to mature and be improved constantly in real-life use. This makes the process of creating a pattern language important. The validation of the KSF was made in industrial case studies. The resulting Knowledge Sharing Pattern Language was validated through justifying the single patterns separately, evaluating the coverage of the language and evaluating the applicability of the Pattern language in an evaluation workshop.