COEVOLUTION PATTERNS TO DETECT AND MANAGE UML DIAGRAMS CHANGES

Abstract

UML diagrams are divided into different perspectives in modelling a problem domain. Preserving coevolution among these diagrams is very crucial so that they can be updated continuously to reflect software changes. Formal methods such as Coloured Petri Nets (CPNs) are widely used in detecting and handling coevolution between software artifacts. Although ample progress has been made, it still remains much work to be done in further improving the effectiveness and accuracy of the state-of-the-art coevolution techniques in managing changes in UML diagrams. In this research, a set of 84 coevolution patterns for supporting coevolution among UML diagrams are proposed to trace the diagrams’ inconsistencies and to determine the change impact incrementally after updating diagrams elements. Coevolution patterns are applied to UML class, object, activity, statechart, and sequence diagrams to cover the different perspectives of UML diagrams. The researcher uses CPNs as a formal language of modelling case study models for the proposed patterns. CPNs tools simulation and monitoring toolboxes are used to validate and monitor the proposed coevolution patterns models and to collect quantitative data about the patterns.