Designed to Spread: Generative Approaches to Enhance Information Diffusion

2025-11-16

AI Summary

Prior work overlooks how content representation affects information diffusion. Method: This paper introduces Diffusion-Optimized Content Generation (DOCG), a novel generative task that automatically produces highly diffusive multimodal content tailored to specific audiences. We propose the first network-structure-free, content-level diffusion evaluation metric and design an interpretable editing framework integrating reinforcement learning with generative models to jointly optimize semantic fidelity and diffusion enhancement. Our approach incorporates audience-aware mechanisms and influence modeling to enable coordinated optimization of textual and visual content. Contribution/Results: Evaluated on real-world social datasets and user studies, DOCG significantly improves diffusion reach (+28.3%) while preserving semantic integrity (semantic similarity ≥ 0.91), demonstrating both effectiveness and practical applicability.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Natural Language Processing: Generation
• Search and Optimization: Learning to Search

Application Category

• Web Mining and Content Analysis: Content-based information diffusion
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• User Modeling, Personalization and Recommendation: Personalized content ranking and presentation

Abstract

Social media has fundamentally transformed how people access information and form social connections, with content expression playing a critical role in driving information diffusion. While prior research has focused largely on network structures and tipping point identification, it provides limited tools for automatically generating content tailored for virality within a specific audience. To fill this gap, we propose the novel task of DOCG and introduce an information enhancement algorithm for generating content optimized for diffusion. Our method includes an influence indicator that enables content-level diffusion assessment without requiring access to network topology, and an information editor that employs reinforcement learning to explore interpretable editing strategies. The editor leverages generative models to produce semantically faithful, audience-aware textual or visual content. Experiments on real-world social media datasets and user study demonstrate that our approach significantly improves diffusion effectiveness while preserving the core semantics of the original content.

Problem

• Automatically generating viral content tailored for specific audiences
• Assessing content-level diffusion without network topology access
• Optimizing information diffusion while preserving original semantics

Innovation

• Generative models create audience-aware content
• Reinforcement learning explores interpretable editing strategies
• Influence indicator assesses diffusion without network topology