Balancing Accuracy and Efficiency in Multi-Turn Intent Classification for LLM-Powered Dialog Systems in Production

TL;DR

This paper introduces novel LLM-based methods, Symbol Tuning and C-LARA, to improve multi-turn intent classification accuracy and efficiency in dialogue systems, addressing data scarcity and latency challenges.

Contribution

It presents two innovative approaches leveraging LLMs for data augmentation and task simplification, enhancing scalability and performance in production dialogue systems.

Findings

Improved intent classification accuracy by 5.09%

Reduced annotation costs by 40%

Enabled scalable deployment in low-resource settings

Abstract

Accurate multi-turn intent classification is essential for advancing conversational AI systems. However, challenges such as the scarcity of comprehensive datasets and the complexity of contextual dependencies across dialogue turns hinder progress. This paper presents two novel approaches leveraging Large Language Models (LLMs) to enhance scalability and reduce latency in production dialogue systems. First, we introduce Symbol Tuning, which simplifies intent labels to reduce task complexity and improve performance in multi-turn dialogues. Second, we propose C-LARA (Consistency-aware, Linguistics Adaptive Retrieval Augmentation), a framework that employs LLMs for data augmentation and pseudo-labeling to generate synthetic multi-turn dialogues. These enriched datasets are used to fine-tune a small, efficient model suitable for deployment. Experiments conducted on multilingual dialogue datasets demonstrate significant improvements in classification accuracy and resource efficiency. Our methods enhance multi-turn intent classification accuracy by 5.09%, reduce annotation costs by 40%, and enable scalable deployment in low-resource multilingual industrial systems, highlighting their practicality and impact.