Inconsistency Management Framework for Model-Based Development

Alexander Reder
Johannes Kepler University, Austria

In contrast to programming environments, model-based development tools lack in an efficient support in detecting and repairing design errors. However, inconsistencies must be resolved eventually and the detection of inconsistencies is of little use if it is not known how to use this information. Quite many approaches exist for detecting inconsistencies. Only some of them provide solutions for repairing individual inconsistencies but none of them are able to investigate the repair problem comprehensively – in particular considering the side effects that might occur when applying a repair. My PhD thesis focuses on resolving inconsistencies, the different strategies one can follow and how to deal with the critical problem of side effects. My approach is based on an incremental approach for detecting inconsistencies and combines runtime analysis of the design rules’ evaluation behavior with static analysis on the design rules’ structure. The main contribution is an efficient and effective inconsistency management framework that can be applied generically to (most) modeling and design rule languages.