LLM-assisted clustering framework improves intent discovery in dialogue systems

Researchers have developed NILC, a new framework designed to improve new intent discovery in dialogue systems. NILC iteratively refines text embeddings and cluster centroids using large language models (LLMs) to better capture nuanced semantic meanings. The framework also incorporates LLM-generated semantic centroids and augments ambiguous utterances to enhance cluster accuracy, showing significant performance gains across various datasets in both unsupervised and semi-supervised settings. AI

IMPACT Enhances dialogue system capabilities by improving the accuracy of recognizing user intents.

RANK_REASON The cluster contains an academic paper detailing a new method for intent discovery using LLMs. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• Hongtao Wang
• LLM

• paper
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →