MORGAN: Modeling Recommender system based on GrAph Neural network

MORGAN (Modeling Recommender system based on GrAph Neural network) provides relevant artefacts during the specification of the domain-specific model by exploiting ML/AI capabilities. In particular:
    • SERVICE 1: recommending SysML attributes and classes given a model under construction (UNIVAQ-VCE-JKU)
    • SERVICE 2: providing the next operation by using historical data collected by Process Mining tool (provided by JKU)

    This repository contains all the needed materials to replicate the experiments presented in the paper. We structured the content as follows:

    • Parsers folder contains the two parsers used to extract information from metamodels and models
    • Morgan folder contains the Python implementation of the tool and the datasets employed in the evaluation

    Repository link in GitHub: https://github.com/MDEGroup/MORGAN External link, opens in new window.