Domain Model Evolutionary Approach Based on Semantic Association

Domain models can provide a problem-oriented architecture for the application domain. But the adaptability of domain models will inevitably descend with the development of domain. Thus, the ability of domain models to vary with the domain changes is vital. Based on an implementation framework of domain goal management repository (DGMR), four intelligent domain model evolutionary algorithms are presented in this paper. They utilize the modeling usage records of domain models as the basis of evolution, which contribute to guide the evolution process objectively. They can recommend the evolutionary proposal to domain engineers to help the evolutionary decision process. In the end, a trial application of this approach is proposed to demonstrate the prospect of our approach.