Towards Variability Management in Bidirectional Model Transformation

Xiao He,Zhenjiang Hu,Yi Liu
DOI: https://doi.org/10.1109/compsac.2017.252
2017-01-01
Abstract:The bidirectional model transformation (BX) comprises a forward transformation get and a backward transformation put. Given that get may be an information-loss transformation, the behavior of put may be uncertain. An uncertain put produces many valid outputs that fit different application scenarios. This paper proposes an approach to variability management in BX to enable put to generate an output model with several variation points that can be configured to adapt this output for different uses. Firstly, this paper proposes a variability metamodel and management framework, which are used to characterize and configure variation points in a transformation result model. Secondly, this paper extends a BX language to specify a BX with variability. Thirdly, this paper presents a BX engine, which can execute a BX with variability and generate a model that contains variation points. Lastly, an evaluation is presented to show the feasibility and scalability of our approach.
What problem does this paper attempt to address?