Winning the Social Media Influence Battle: Uncertainty-Aware Opinions to Understand and Spread True Information via Competitive Influence Maximization

TL;DR

This paper introduces a novel DRL framework incorporating Subjective Logic to better model uncertain opinions and user behaviors in competitive influence maximization, effectively spreading true information and mitigating falsehoods in social networks.

Contribution

It presents a new DRL-based approach that integrates uncertainty modeling and strategic competition to improve influence maximization in social networks.

Findings

Outperforms state-of-the-art methods by over 20% in influence spread.

Achieves faster influence propagation under realistic network conditions.

Doubles performance in partially observable networks with disbelieving users.

Abstract

Competitive Influence Maximization (CIM) involves entities competing to maximize influence in online social networks (OSNs). Current Deep Reinforcement Learning (DRL) methods in CIM rely on simplistic binary opinion models (i.e., an opinion is represented by either 0 or 1) and often overlook the complexity of users' behavioral characteristics and their prior knowledge. We propose a novel DRL-based framework that enhances CIM analysis by integrating Subjective Logic (SL) to accommodate uncertain opinions, users' behaviors, and their preferences. This approach targets the mitigation of false information by effectively propagating true information. By modeling two competitive agents, one spreading true information and the other spreading false information, we capture the strategic interplay essential to CIM. Our framework utilizes an uncertainty-based opinion model (UOM) to assess the impact on information quality in OSNs, emphasizing the importance of user behavior alongside network topology in selecting influential seed nodes. Extensive experiments demonstrate that our approach significantly outperforms state-of-the-art methods, achieving faster and more influential results (i.e., outperforming over 20%) under realistic network conditions. Moreover, our method shows robust performance in partially observable networks, effectively doubling the performance when users are predisposed to disbelieve true information.