Towards a hybrid semi-automatic technique for metamodel matching

During the last two decades the theory of model transformation has been extensively studied and subject of many research. So far, several transformation languages have been developed and implemented. Recently, a metamodel matching technique has been proposed in order to find mapping models that support the deduction of transformations definitions. The model matching aims at searching semantic similarity between models elements, and therefore, facilitating semi-automatic mapping generation. However, current model matching techniques mostly focus on some comparison criteria (i.e., pre-matching effort, input model, output model, quality measures values) while ignore others; in addition these techniques are used in isolation, so that their advantages and commons are not exploited. In this paper, we tackle these issues by proposing a new technique relying on combination of metamodels matching techniques. To reach our goal, schema and ontology matching techniques can be adapted to develop a hybrid technique for metamodel matching. This technique will produce better matching results increasing the mappings quality.