A Variability-Driven Analysis Method for Automatic Extraction of Domain Behaviors

Reinhartz-Berger, Iris; Abbas, Sameh

Domain engineering focuses on modeling knowledge in an application domain for supporting systematic reuse in the context of complex and constantly evolving systems. Automatically supporting this task is challenging; most existing methods assume high similarity of variants which limits reuse of the generated domain artifacts, or provide very low-level features rather than actual domain features. As a result, these methods are limited in handling common scenarios such as similarly behaving systems developed by different teams, or merging existing products. To address this gap, we propose a method for extracting domain knowledge in the form of domain behaviors, building on a previously developed framework for behavior-based variability analysis among class operations. Machine learning techniques are applied for identifying clusters of operations that can potentially form domain behaviors. The approach is evaluated on a set of open-source video games, named apo-games.