MM-OpenFGL: A Comprehensive Benchmark for Multimodal Federated Graph Learning

TL;DR

This paper introduces MM-OpenFGL, a comprehensive benchmark for multimodal federated graph learning, providing datasets, simulation strategies, and evaluations to advance research in privacy-preserving, multimodal graph analysis.

Contribution

It is the first systematic benchmark for MMFGL, formalizing the paradigm and enabling rigorous evaluation with diverse datasets, tasks, and methods.

Findings

MM-OpenFGL includes 19 datasets across 7 domains.

Extensive experiments reveal insights into effectiveness and robustness of MMFGL methods.

The benchmark facilitates future research and development in multimodal federated graph learning.

Abstract

Multimodal-attributed graphs (MMAGs) provide a unified framework for modeling complex relational data by integrating heterogeneous modalities with graph structures. While centralized learning has shown promising performance, MMAGs in real-world applications are often distributed across isolated platforms and cannot be shared due to privacy concerns or commercial constraints. Federated graph learning (FGL) offers a natural solution for collaborative training under such settings; however, existing studies largely focus on single-modality graphs and do not adequately address the challenges unique to multimodal federated graph learning (MMFGL). To bridge this gap, we present MM-OpenFGL, the first comprehensive benchmark that systematically formalizes the MMFGL paradigm and enables rigorous evaluation. MM-OpenFGL comprises 19 multimodal datasets spanning 7 application domains, 8 simulation strategies capturing modality and topology variations, 6 downstream tasks, and 57 state-of-the-art methods implemented through a modular API. Extensive experiments investigate MMFGL from the perspectives of necessity, effectiveness, robustness, and efficiency, offering valuable insights for future research on MMFGL.