Using Critiquing Systems for Inconsistency Detection in Software Engineering Models

Many approaches have been proposed for consistency management of software engineering documents and specifications. A few others have been proposed to check consistency among software engineering models. For example, abstract state machines, knowledge-based approaches and so on. In this paper, we apply a different technique that uses critiquing systems. A critiquing system monitors user's actions and triggers a signal when one of those actions activates pre-specified rules, called critics. Because critics are small, we argue that they might be used to address two open issues in inconsistency detection, namely efficiency and scalability. An example of this approach is presented to check domain engineering models (feature diagrams) and application engineering models (class diagrams). Feature diagrams are used to abstractly and concisely express commonality and variability across a domain. These diagrams are used as source of information in the generation of critics in UML class diagrams. We present an environment, called DAISY that uses three different critiquing systems to demonstrate the feasibility of our approach.