RDB2G-Bench: A Comprehensive Benchmark for Automatic Graph Modeling of Relational Databases

2025-06-02

AI Summary

Diverse relational database (RDB)-to-graph modeling approaches exist, yet the absence of a systematic evaluation benchmark impedes identification of optimal graph structures. Method: We introduce the first comprehensive benchmark for automatic RDB-to-graph modeling, establishing a standardized evaluation framework that encompasses five real-world databases, twelve downstream prediction tasks, and approximately 50,000 graph-structure–performance pairs. All graphs are uniformly constructed using foreign-key relationships to define nodes and edges; nine state-of-the-art automated graph construction methods are integrated, and graph structures are precomputed to ensure consistent training and evaluation. Contribution/Results: Our framework accelerates evaluation by 600× compared to on-the-fly graph construction. It empirically identifies critical structural patterns governing modeling efficacy—revealing, for the first time, actionable insights into optimal graph design—and provides a reproducible, evidence-based standard for RDB-to-graph modeling assessment.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion
• Knowledge Representation and Reasoning: Computational Complexity of Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Relational databases (RDBs) are composed of interconnected tables, where relationships between them are defined through foreign keys. Recent research on applying machine learning to RDBs has explored graph-based representations of RDBs, where rows of tables are modeled as nodes, and foreign key relationships are modeled as edges. RDB-to-graph modeling helps capture cross-table dependencies, ultimately leading to enhanced performance across diverse tasks. However, there are numerous ways to model RDBs as graphs, and performance varies significantly depending on the chosen graph model. In our analysis, applying a common heuristic rule for graph modeling leads to up to a 10% drop in performance compared to the best-performing graph model, which remains non-trivial to identify. To foster research on intelligent RDB-to-graph modeling, we introduce RDB2G-Bench, the first benchmark framework for evaluating such methods. We construct extensive datasets covering 5 real-world RDBs and 12 predictive tasks, resulting in around 50k graph-performance pairs for efficient and reproducible evaluations. Thanks to our precomputed datasets, we were able to benchmark 9 automatic RDB-to-graph modeling methods on the 12 tasks over 600x faster than on-the-fly evaluation, which requires repeated model training. Our analysis of the datasets and benchmark results reveals key structural patterns affecting graph model effectiveness, along with practical implications for effective graph modeling.

Problem

• Evaluating performance of various RDB-to-graph modeling methods
• Identifying optimal graph models for relational databases
• Providing benchmark framework for intelligent RDB-to-graph conversion

Innovation

• Benchmark framework for RDB-to-graph modeling
• Precomputed datasets for efficient evaluations
• Analysis of structural patterns in graph models