Abstract

The presence of crosscutting concerns in legacy systems makes the maintenance an error-prone and time-consuming activity. Architecture-Driven Modernization (ADM) is a good alternative for modernizing systems in a modeldriven way by means of Knowledge Discovery Metamodel (KDM). However, the current proposal and tools did not take into account mining of crosscutting concerns. We present CCKDM, a tool for identifying crosscutting concerns which is a combination of a concern library and a modified clustering algorithm. The input is a KDM model and the output is the same KDM with the identified concerns annotated. The last one would enable to build and apply concern-based refactorings in order to have target KDM with the concerns better modularized.