Towards clone detection in UML domain models

  • Authors:
  • Harald Störrle

  • Affiliations:
  • Technical University of Denmark, Lyngby, Denmark

  • Venue:
  • Proceedings of the Fourth European Conference on Software Architecture: Companion Volume
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

Code clones are a substantial problem for code based development; it is not surprising that model clones are a problem for model based development, too. Past experience with model version control vs. code version control suggests, however, that there are radically different problems and techniques on the code level than on the model level. In this paper, we explore the problems and possibilities associated with detecting clones in UML domain models. In particular, we have designed a number of algorithms and heuristics to detect clones, and have implemented them in the MQlone tool (read as "m clone"). However, all of the approaches we have studied so far have their weak spots, so that we cannot present a universal solution. Further work is required to refine and combine the approaches studied here, and test them on larger populations of models. While the related work faces similar unsolved problems our approach is the first one to deal with UML as a whole rather than with Matlab/Simulink or individual UML model types.