Stance Classification

Stance classification aims to automatically determine the viewpoint of an author or speaker towards a given topic, a crucial task with applications in areas like misinformation detection and social media analysis. Current research focuses on leveraging large language models (LLMs), often enhanced with techniques like chain-of-thought prompting and fine-tuning, as well as graph neural networks to incorporate network information from social media interactions. These advancements are improving accuracy, particularly in handling nuanced language and implicit stances, but challenges remain in addressing biases within LLMs and handling the temporal evolution of stances.