Towards a Model-driven Approach for Context-aware Software

Context-aware software (C-AS) uses context to provide users with relevant information or services. Although C-AS has an increasing importance, this area is still immature in terms of tools, languages, and methods. Therefore, C-AS presents several opportunities and challenges for software engineers, and new models and theories are needed to address these challenges. Some of the main problems in software engineering for C-AS are the lack of mechanisms for representing some relevant concepts in the requirements and design specifications directly, and the complexity to translate requirements and design specifications to a concrete implementation. As an alternative of solution, the research project proposes MiDAS, a framework that uses a model-driven approach to develop C-AS. Specifically, MiDAS comprises two domain-specific languages to specify requirements and design in C-AS; a model-to-model transformation to derive design models from requirements models; a modelto-text transformation to derive a platform independent model (PIM) from design models; and a model-to-text transformation to derive context-aware software for specific platforms, taking as input the PIM.