Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Large Language Models Meet Stance Detection: A Survey of Tasks, Methods, Applications, Challenges and Future Directions

Lata Pangtey, Anukriti Bhatnagar, Shubhi Bansal, Shahid Shafi Dar, Nagendra Kumar

TL;DR

The paper surveys how large language models transform stance detection across textual and multimodal domains, introducing a taxonomy along learning methods, data modalities, and target relationships. It synthesizes datasets, benchmarks, and performance trends, and discusses agentic and tool-augmented LLMs, with applications spanning misinformation, political discourse, and public health. Key contributions include a comprehensive taxonomy, cross-domain and cross-lingual analyses, and a synthesis of challenges such as implicit stance expression, bias, and computational demands, alongside future directions like explainable reasoning and real-time deployment. The work provides a roadmap for researchers and practitioners to develop next-generation stance detection systems powered by LLMs, with emphasis on multimodal fusion, cross-lingual transfer, and integration with fact-checking and moderation frameworks.

Abstract

Stance detection is essential for understanding subjective content across various platforms such as social media, news articles, and online reviews. Recent advances in Large Language Models (LLMs) have revolutionized stance detection by introducing novel capabilities in contextual understanding, cross-domain generalization, and multimodal analysis. Despite these progressions, existing surveys often lack comprehensive coverage of approaches that specifically leverage LLMs for stance detection. To bridge this critical gap, our review article conducts a systematic analysis of stance detection, comprehensively examining recent advancements of LLMs transforming the field, including foundational concepts, methodologies, datasets, applications, and emerging challenges. We present a novel taxonomy for LLM-based stance detection approaches, structured along three key dimensions: 1) learning methods, including supervised, unsupervised, few-shot, and zero-shot; 2) data modalities, such as unimodal, multimodal, and hybrid; and 3) target relationships, encompassing in-target, cross-target, and multi-target scenarios. Furthermore, we discuss the evaluation techniques and analyze benchmark datasets and performance trends, highlighting the strengths and limitations of different architectures. Key applications in misinformation detection, political analysis, public health monitoring, and social media moderation are discussed. Finally, we identify critical challenges such as implicit stance expression, cultural biases, and computational constraints, while outlining promising future directions, including explainable stance reasoning, low-resource adaptation, and real-time deployment frameworks. Our survey highlights emerging trends, open challenges, and future directions to guide researchers and practitioners in developing next-generation stance detection systems powered by large language models.

Large Language Models Meet Stance Detection: A Survey of Tasks, Methods, Applications, Challenges and Future Directions

TL;DR

The paper surveys how large language models transform stance detection across textual and multimodal domains, introducing a taxonomy along learning methods, data modalities, and target relationships. It synthesizes datasets, benchmarks, and performance trends, and discusses agentic and tool-augmented LLMs, with applications spanning misinformation, political discourse, and public health. Key contributions include a comprehensive taxonomy, cross-domain and cross-lingual analyses, and a synthesis of challenges such as implicit stance expression, bias, and computational demands, alongside future directions like explainable reasoning and real-time deployment. The work provides a roadmap for researchers and practitioners to develop next-generation stance detection systems powered by LLMs, with emphasis on multimodal fusion, cross-lingual transfer, and integration with fact-checking and moderation frameworks.

Abstract

Stance detection is essential for understanding subjective content across various platforms such as social media, news articles, and online reviews. Recent advances in Large Language Models (LLMs) have revolutionized stance detection by introducing novel capabilities in contextual understanding, cross-domain generalization, and multimodal analysis. Despite these progressions, existing surveys often lack comprehensive coverage of approaches that specifically leverage LLMs for stance detection. To bridge this critical gap, our review article conducts a systematic analysis of stance detection, comprehensively examining recent advancements of LLMs transforming the field, including foundational concepts, methodologies, datasets, applications, and emerging challenges. We present a novel taxonomy for LLM-based stance detection approaches, structured along three key dimensions: 1) learning methods, including supervised, unsupervised, few-shot, and zero-shot; 2) data modalities, such as unimodal, multimodal, and hybrid; and 3) target relationships, encompassing in-target, cross-target, and multi-target scenarios. Furthermore, we discuss the evaluation techniques and analyze benchmark datasets and performance trends, highlighting the strengths and limitations of different architectures. Key applications in misinformation detection, political analysis, public health monitoring, and social media moderation are discussed. Finally, we identify critical challenges such as implicit stance expression, cultural biases, and computational constraints, while outlining promising future directions, including explainable stance reasoning, low-resource adaptation, and real-time deployment frameworks. Our survey highlights emerging trends, open challenges, and future directions to guide researchers and practitioners in developing next-generation stance detection systems powered by large language models.
Paper Structure (42 sections, 7 equations, 10 figures, 3 tables)

This paper contains 42 sections, 7 equations, 10 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Large Language Models: A Brief Overview
  3. Evolution and Architectures
  4. Pretraining and Fine-tuning Paradigms
  5. In-context Learning, Prompt Engineering, and Instruction Tuning
  6. Performance in Downstream Tasks
  7. Learning Methods
  8. Supervised Approach
  9. Unsupervised Approach
  10. Zero-Shot Learning
  11. Few-Shot Learning
  12. Approaches
  13. Unimodal Approaches
  14. Multimodal Approaches
  15. Multilingual Approaches
  16. ...and 27 more sections

Figures (10)

  • Figure 1: General Framework for Stance Detection
  • Figure 2: Term Frequency Word Cloud Summarizing Prominent Vocabulary in Stance Detection Research
  • Figure 5: Outline of the Survey on Stance Detection
  • Figure 6: Evolution of Language Models: From Rule-Based Systems to Modern Large Language Models (LLMs)
  • Figure 7: Downstream Applications of LLMs
  • ...and 5 more figures