Semantic consistency checking for model transformations

Model transformation, as a key technique of MDA, is error-prone because of conceptual flaws in design and man-made errors in manual transformation rules. So the consistency checking of model transformations is of great importance for MDA. In this paper, a framework of semantic consistency checking for model transformation is proposed and discussed. In this framework, a graph representation is required to describe model languages, model transformation rules, and source code. Then several semantic properties are selected to be studied, and algorithms based on critical pairs are given to check whether these properties are preserved by model transformations. At last, a case study is performed to demonstrate the feasibility.