Dynamic Label Name Refinement for Few-Shot Dialogue Intent Classification

TL;DR

This paper introduces a dynamic label refinement method using large language models to improve few-shot dialogue intent classification, effectively reducing confusion among similar intents and enhancing interpretability.

Contribution

It presents a novel in-context learning approach with dynamic label refinement to better distinguish intents in few-shot settings, addressing semantic overlap issues.

Findings

Significantly improves intent classification accuracy across datasets.

Reduces confusion between semantically similar intents.

Produces more interpretable and semantically coherent intent labels.

Abstract

Dialogue intent classification aims to identify the underlying purpose or intent of a user's input in a conversation. Current intent classification systems encounter considerable challenges, primarily due to the vast number of possible intents and the significant semantic overlap among similar intent classes. In this paper, we propose a novel approach to few-shot dialogue intent classification through in-context learning, incorporating dynamic label refinement to address these challenges. Our method retrieves relevant examples for a test input from the training set and leverages a large language model to dynamically refine intent labels based on semantic understanding, ensuring that intents are clearly distinguishable from one another. Experimental results demonstrate that our approach effectively resolves confusion between semantically similar intents, resulting in significantly enhanced performance across multiple datasets compared to baselines. We also show that our method generates more interpretable intent labels, and has a better semantic coherence in capturing underlying user intents compared to baselines.