Guidelines and metrics for configurable and sustainable architectural knowledge modelling

Architectural Knowledge Management (AKM) has been an active research area in the last decade; the importance of making the right architectural decisions -- and making these at the right time -- has been recognized by the contemporary software engineering practices. Several AKM meta-models, templates and tools have been proposed and applied in practice to capture architectural design decisions and minimize architectural drift during software evolution. However, most of these AKM models, and the architectural decisions captured with them, lack contextual awareness, flexibility and maintainability over time. In this position paper, we outline an extended AKM meta-model and a set of guidelines with the goal to (i) allow AKM tool engineers to construct more configurable and therefore flexible AKM tools, (ii) allow knowledge engineers and method coaches to create more sustainable and therefore maintainable decision logs (AK model instances). We approach these two goals by way of mapping the extended AKM meta-model concepts to quality attributes for architectural knowledge as well as supporting AK metrics