Chain of Stance: Stance Detection with Large Language Models

TL;DR

This paper introduces Chain of Stance, a novel prompting method for large language models that decomposes stance detection into intermediate steps, significantly improving performance on benchmark datasets.

Contribution

It proposes a new prompting approach that leverages LLMs as expert stance detectors through intermediate assertions, outperforming existing fine-tuning methods.

Findings

Achieves state-of-the-art F1 score of 79.84 in few-shot stance detection.

Demonstrates significant performance gains over existing methods.

Effective in zero-shot and few-shot learning scenarios.

Abstract

Stance detection is an active task in natural language processing (NLP) that aims to identify the author's stance towards a particular target within a text. Given the remarkable language understanding capabilities and encyclopedic prior knowledge of large language models (LLMs), how to explore the potential of LLMs in stance detection has received significant attention. Unlike existing LLM-based approaches that focus solely on fine-tuning with large-scale datasets, we propose a new prompting method, called \textit{Chain of Stance} (CoS). In particular, it positions LLMs as expert stance detectors by decomposing the stance detection process into a series of intermediate, stance-related assertions that culminate in the final judgment. This approach leads to significant improvements in classification performance. We conducted extensive experiments using four SOTA LLMs on the SemEval 2016 dataset, covering the zero-shot and few-shot learning setups. The results indicate that the proposed method achieves state-of-the-art results with an F1 score of 79.84 in the few-shot setting.