MT2-CSD: A New Dataset and Multi-Semantic Knowledge Fusion Method for Conversational Stance Detection

2025-06-26

AI Summary

Existing stance detection research is largely confined to single-instance settings and suffers from a lack of realistic multi-turn interactive data. To address this, this paper proposes a novel paradigm for stance detection in social media multi-target dialogues. First, we introduce MT2-CSD—the largest publicly available annotated dataset to date (24,457 instances)—which uniquely supports multi-turn, multi-participant, and multi-issue stance modeling. Second, we propose LLM-CRAN, a Large Language Model-enhanced Conversation Relation Attention Network that explicitly captures dialogue logical structure, inter-utterance relational dependencies, and multi-granularity semantic knowledge for end-to-end stance identification. Experiments on MT2-CSD demonstrate that LLM-CRAN significantly outperforms diverse strong baselines, validating its effectiveness and state-of-the-art performance in complex, realistic dialogue scenarios.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

In the realm of contemporary social media, automatic stance detection is pivotal for opinion mining, as it synthesizes and examines user perspectives on contentious topics to uncover prevailing trends and sentiments. Traditional stance detection research often targets individual instances, thereby limiting its capacity to model multi-party discussions typical in real social media scenarios. This shortcoming largely stems from the scarcity of datasets that authentically capture the dynamics of social media interactions, hindering advancements in conversational stance detection. In this paper, we introduce MT2-CSD, a comprehensive dataset for multi-target, multi-turn conversational stance detection. To the best of our knowledge, MT2-CSD is the largest dataset available for this purpose, comprising 24,457 annotated instances and exhibiting the greatest conversational depth, thereby presenting new challenges for stance detection. To address these challenges, we propose the Large Language model enhanced Conversational Relational Attention Network (LLM-CRAN), which exploits the reasoning capabilities of LLMs to improve conversational understanding. We conduct extensive experiments to evaluate the efficacy of LLM-CRAN on the MT2-CSD dataset. The experimental results indicate that LLM-CRAN significantly outperforms strong baseline models in the task of conversational stance detection.

Problem

• Lack of datasets for multi-party social media stance detection
• Need for improved conversational understanding in stance detection
• Challenges in modeling multi-target, multi-turn conversational dynamics

Innovation

• Introduces MT2-CSD dataset for stance detection
• Proposes LLM-CRAN for conversational understanding
• Leverages LLMs to enhance relational attention