Connecting the Models: A Global Mega-model of MDE Projects on GitHub

2020-12-17 · Source: cs.SE updates on arXiv.org · Field: Technology & Digital — Software Development & Engineering, Data Science & Analytics · Depth: Expert, extended

Summary

A new study introduces ModelGraph, a global mega-model and dataset derived from 7,436 GitHub projects containing over 325,000 Model-Driven Engineering (MDE) artefacts. Published in March 2026, this resource details relationships between MDE artefacts like Ecore, ATL, and Xtext. The methodology involved mining GitHub, recovering project-specific mega-models, and constructing a global mega-model by analyzing near-duplicates. This comprehensive dataset, available with a web-based exploration tool, aims to foster empirical analysis of MDE tool usage and inter-project dependencies. It addresses a long-standing gap in understanding MDE in practice.

Key takeaway

For Research Scientists studying Model-Driven Engineering, this ModelGraph dataset offers an unprecedented resource to empirically analyze MDE tool usage and inter-project dependencies. You can explore technology combinations, identify reuse patterns, and train specialized AI models, overcoming previous data limitations. Utilize the provided web tool and APIs to investigate MDE project maturity and evolution.

Key insights

A global mega-model and dataset reveals MDE artefact relationships and reuse patterns across 7,436 GitHub projects.

Principles

• MDE lacks package managers, leading to copy-paste reuse.
• Implicit artefact dependencies require specialized recovery methods.

Method

Mine GitHub for MDE artefacts, recover project-level mega-models by analyzing dependencies, then merge into a global mega-model via near-duplicate detection.

In practice

• Explore MDE project health and technology combinations.
• Train AI models using deduplicated MDE artefacts.
• Identify central hubs and structural patterns.

Topics

• Model-Driven Engineering
• GitHub Mining
• Mega-models
• EMF Ecosystem
• Artefact Duplication
• Empirical Software Engineering

Code references

• utwente-fmt/attop
• upohl/mechatronicuml
• fraunhofer-iem/uppaal-model
• fraunhofer-iem/mechatronicuml
• NCIP/lexevs

Best for: AI Scientist, Research Scientist, Software Engineer

Related on AIssential

• Mila World Modeling Workshop: wrap-up — Advancing AI requires systems that perceive, reason, and act in the real world, integrating safety and scientific understanding.
• Netflix Introduces ‘Model Lifecycle Graph’ to Scale Enterprise Machine Learning — Graph-based architectures can effectively manage complex ML system dependencies and improve discoverability at enterprise scale.
• It's not the Language Model, it's the Tool: Deterministic Mediation for Scientific Workflows — Deterministic tools orchestrated by language models ensure reproducibility in scientific workflows, unlike direct code generation.
• 10 GitHub Repositories to Master System Design — GitHub repositories offer structured learning paths for mastering system design fundamentals and specialized architectures.
• An open-source image variation dataset (Apache 2.0) — A new open-source image variation dataset is released, crucial for training the sub-10B Lunara model.
• A Cartography of Open Collaboration in Open Source AI: Mapping Practices, Motivations, and Governance in 14 Open Large Language Model Projects — Open collaboration in LLM development is multifaceted, driven by diverse motivations and structured by varied governance models.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by cs.SE updates on arXiv.org.