Fair Densest Subgraph Across Multiple Graphs

2025-02-03

ECML/PKDD

AI Summary

This work addresses the problem of dense subgraph mining across multiple graphs, proposing the first joint optimization framework incorporating group fairness constraints to simultaneously identify high-density, cross-graph consistent, and group-balanced common subgraphs. Methodologically, we formulate an integer programming model that integrates graph kernel alignment with fairness regularization, and develop a provably approximate solution via Lagrangian relaxation and iterative projection. Our key contribution is the formal incorporation of group fairness into multi-graph dense subgraph discovery—breaking from conventional density-only optimization paradigms. Experiments on multi-social-network and biological interaction graphs demonstrate that our method achieves substantial gains in cross-graph consistency while maintaining overall density above a prescribed threshold: density loss remains under 5%, consistency improves significantly, and group fairness metrics increase by 3.2× over state-of-the-art baselines.

Technology Category

• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Machine Learning: Graph-based Machine Learning

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Problem

• Graph Analysis
• Dense Region Detection
• Consistency Evaluation

Innovation

• Dense Region Detection
• Multi-Graph Analysis
• Approximation Algorithm