Automated Graph Machine Learning Operations (MLOps) Workflow

This thesis introduces Auto-GMLOps, an automated workflow for Graph Machine Learning (ML) operations, addressing challenges in graph data engineering, automated GNN model design, and deployment. By streamlining these processes, it enables more efficient utilization of graph data and the development of tailored GNN models. Practical applications include fraud detection in financial transactions, disease prediction in healthcare networks, recommendation systems in social networks, and traffic pattern analysis for urban planning.