Towards efficient loading of change-based models

Abstract

This paper proposes and evaluates an efficient approach for loading models stored in a change-based format. The work builds on language-independent change-based persistence (CBP) of models conforming to object-oriented metamodelling architectures such as MOF and EMF, an approach which persists a model’s editing history rather than its current state. We evaluate the performance of the proposed loading approach and assess its impact on saving change-based models. Our results show that the proposed approach significantly improves loading times compared to the baseline CBP loading approach, and has a negligible impact on saving.