Enhanced Dependency Structure Matrix for Moose

International audienceDependency Structure Matrix (DSM), an approach de- veloped in the context of process optimization, has been suc- cessfully applied to identify software dependencies among packages and subsystems. It exists a couple of algorithms to help organizing the matrix in a form that reflects the ar- chitecture and highlights patterns and problematic depen- dencies between subsystems. However, the existing DSM implementations often miss some important information in their visualization to fully support a reengineering effort. In this paper we enhanced DSM with enriched cell contextual information by showing information (i) about the kinds of references made (inheritance, class accesses..), (ii) the pro- portion of entities (classes/methods) doing references, (iii) the proportion of entities been the target of the references. We distinguish independent cycles and stress the cycles us- ing coloring information. This work has been implemented on top of the Moose open-source reengineering environment and Mondrian. It has been applied to non-trivial case stud- ies such as the Morphic UI frameworks available in Squeak an open-source Smalltalk