Model-Driven Engineering (MDE) approach to generate Machine Learning Code based on SysML and Mapping Configuration
This repository contains code for the Proof-of-concept (PoC) implementation done for the Master of Science in Information and Software Engineering at the University of Applied Sciences Dornbirn.
This work was supervised and conceptually elaborated at the Chair of Information Systems and Business Process Management (i17), Department of Computer Science, Technical University of Munich
This PoC implementation is based on a recently published approach to machine learning modeling with SysML at INDIN 2022 (Radler, S., Rigger, E., Mangler, J., Rinderle-Ma, S., 2022. Integration of Machine Learning Task Definition in Model-Based Systems Engineering using SysML, in: 2022 IEEE 20th International Conference on Industrial Informatics (INDIN). Presented at the 2022 IEEE 20th International Conference on Industrial Informatics (INDIN), IEEE, Perth, Australia, pp. 546–551. https://doi.org/10.1109/INDIN51773.2022.9976107).
The goals of the PoC implementation are as follows:
- Use of an existing SysML profile for machine learning (ML) modeling, which can be used to describe data, interfaces, ML processing, and to define ML algorithms
- Develop a model transformation to extract information from the SysML model into the intermediate model. The intermediate model must be defined as a metamodel beforehand
- Generating executable ML code from a SysML model using the intermediate model and template-based code generation
- Definition of ML templates for specific stereotypes from the user-defined SysML profile, which are used as the basis for generation
- Provide a mapping configuration mechanism that allows the user to map SysML elements to template variables and exchange them dynamically
MODELS 2023 TODO right after publication is done.