MoQuaVas - Models for High Quality and Value Based Software Engineering

The project focuses on four aspects: Quality Models with their hierarchical structure allow capturing the knowledge about software quality in a comprehensive way. Large quality models – as built and maintained by Siemens in the last years – comprise thousands of rules and a large number of classifying quality attributes. Tailoring of such models is difficult and cumbersome as selecting the needed quality attributes, metrics and tools requires good knowledge of the application domain. Even more challenging is the maintenance of the quality model, with new perspectives on quality to be considered (e.g., ISO/IEC 5055) and new metrics and rules that have to be consistently integrated. For this purpose, we started developing a semantic model and associated toolset that currently already allows storing and tailoring quality models based on RDF. Development of a self-assessment system for agile projects that allows to plan, execute and monitor assessments by means of a web-based self-assessment system. The major quality attribute to be fulfilled is the security and especially privacy of data. The design of the system must ensure, that tracing assessment data back to the individuum is not possible by design. Broad literature review to identify and classify practices, smells and methods that help to build good models in computer science and related fields. Terms associated with such approaches are semantic modelling, ontologies or conceptual modelling. Development and validation of tool support for value prioritization based on existing value prioritization methods. The tool support must be integrated with GitLab or other ALM tools like Polarion or TFS to allow for a continuous planning and re-planning of features and epics based on project and operational data. If applicable, standards like OSLC must be considered to avoid vendor lock-in with specific ALM tools.