REM: A Scalable Reinforced Multi-Expert Framework for Multiplex Influence Maximization

2025-01-01

AI Summary

To address the weak generalizability, poor scalability, and heavy reliance on high-quality labeled data in influence maximization for multilayer social networks, this paper proposes a reinforcement learning–based collaborative multi-expert framework. We innovatively formulate seed selection as a learnable policy, integrating graph neural networks with a Mixture-of-Experts (MoE) architecture to dynamically model heterogeneous diffusion processes across layers—enabling fully unsupervised, autonomous optimization of seed sets. Our method jointly leverages policy gradient reinforcement learning and multilayer diffusion modeling. Extensive experiments on multiple real-world datasets demonstrate significant improvements over state-of-the-art methods: up to 18.7% higher influence spread, 3.2× faster inference speed, and scalable support for networks with over one million nodes.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Search and Optimization: Learning to Search
• Multiagent Systems: Multiagent Learning

Application Category

• Social Networks and Social Media: Influence propagation, information diffusion, and the prediction on networks
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Content-based information diffusion

Abstract

In social online platforms, identifying influential seed users to maximize influence spread is a crucial as it can greatly diminish the cost and efforts required for information dissemination. While effective, traditional methods for Multiplex Influence Maximization (MIM) have reached their performance limits, prompting the emergence of learning-based approaches. These novel methods aim for better generalization and scalability for more sizable graphs but face significant challenges, such as (1) inability to handle unknown diffusion patterns and (2) reliance on high-quality training samples. To address these issues, we propose the Reinforced Expert Maximization framework (REM). REM leverages a Propagation Mixture of Experts technique to encode dynamic propagation of large multiplex networks effectively in order to generate enhanced influence propagation. Noticeably, REM treats a generative model as a policy to autonomously generate different seed sets and learn how to improve them from a Reinforcement Learning perspective. Extensive experiments on several real-world datasets demonstrate that REM surpasses state-of-the-art methods in terms of influence spread, scalability, and inference time in influence maximization tasks.

Problem

• Influence Identification
• Information Diffusion
• Learning Sample Scarcity

Innovation

• REM Framework
• Hybrid Technique
• Self-optimized Influencer Selection