Verifying Rumors via Stance-Aware Structural Modeling

2025-12-15

AI Summary

Social media rumor verification faces challenges in jointly modeling semantic content, stance, and conversational structure—particularly constrained by the sequence-length limitations of Transformer architectures. To address this, we propose a stance-aware structured modeling framework: (1) a stance encoder that explicitly models the stance distribution across replies; (2) stance-category–aware reply embedding aggregation; and (3) integration of structural covariates—including hierarchical depth—to alleviate long-sequence modeling bottlenecks. Our approach is the first to co-encode stance signals and conversational topology within a unified architecture. Extensive experiments on multiple benchmark datasets demonstrate significant improvements over state-of-the-art methods in verification accuracy, early detection capability, and cross-platform generalization.

Technology Category

• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)
• Machine Learning: Structured Learning
• Knowledge Representation and Reasoning: Reasoning with Beliefs

Application Category

• Web Mining and Content Analysis: Bridging structured and unstructured data
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Verifying rumors on social media is critical for mitigating the spread of false information. The stances of conversation replies often provide important cues to determine a rumor's veracity. However, existing models struggle to jointly capture semantic content, stance information, and conversation strructure, especially under the sequence length constraints of transformer-based encoders. In this work, we propose a stance-aware structural modeling that encodes each post in a discourse with its stance signal and aggregates reply embedddings by stance category enabling a scalable and semantically enriched representation of the entire thread. To enhance structural awareness, we introduce stance distribution and hierarchical depth as covariates, capturing stance imbalance and the influence of reply depth. Extensive experiments on benchmark datasets demonstrate that our approach significantly outperforms prior methods in the ability to predict truthfulness of a rumor. We also demonstrate that our model is versatile for early detection and cross-platfrom generalization.

Problem

• Verifying rumors on social media using conversation stances
• Capturing semantic content, stance, and conversation structure jointly
• Enhancing rumor truthfulness prediction with scalable structural modeling

Innovation

• Stance-aware structural modeling for rumor verification
• Aggregates reply embeddings by stance category
• Introduces stance distribution and hierarchical depth covariates