EDDA: A Encoder-Decoder Data Augmentation Framework for Zero-Shot Stance Detection

TL;DR

This paper introduces EDDA, a novel data augmentation framework for zero-shot stance detection that uses large language models and rationales to generate diverse, logically connected training samples, significantly improving performance.

Contribution

The paper presents a new encoder-decoder approach that enhances zero-shot stance detection by generating semantically relevant and syntactically diverse data with rationales, addressing limitations of previous methods.

Findings

Significant performance improvement over state-of-the-art ZSSD methods.

Enhanced semantic relevance and syntactic diversity in augmented data.

Effective utilization of rationales for interpretable learning.

Abstract

Stance detection aims to determine the attitude expressed in text towards a given target. Zero-shot stance detection (ZSSD) has emerged to classify stances towards unseen targets during inference. Recent data augmentation techniques for ZSSD increase transferable knowledge between targets through text or target augmentation. However, these methods exhibit limitations. Target augmentation lacks logical connections between generated targets and source text, while text augmentation relies solely on training data, resulting in insufficient generalization. To address these issues, we propose an encoder-decoder data augmentation (EDDA) framework. The encoder leverages large language models and chain-of-thought prompting to summarize texts into target-specific if-then rationales, establishing logical relationships. The decoder generates new samples based on these expressions using a semantic correlation word replacement strategy to increase syntactic diversity. We also analyze the generated expressions to develop a rationale-enhanced network that fully utilizes the augmented data. Experiments on benchmark datasets demonstrate our approach substantially improves over state-of-the-art ZSSD techniques. The proposed EDDA framework increases semantic relevance and syntactic variety in augmented texts while enabling interpretable rationale-based learning.