ECLM: Entity Level Language Model for Spoken Language Understanding with Chain of Intent

TL;DR

ECLM introduces an entity-level approach with a chain of intent mechanism to improve spoken language understanding by reformulating slot-filling as entity recognition, significantly outperforming existing methods.

Contribution

The paper proposes ECLM, a novel framework that reformulates SLU as entity recognition and introduces a chain of intent concept for multi-intent recognition, enhancing performance over baselines.

Findings

ECLM outperforms baselines by 3.7% on MixATIS and 3.1% on MixSNIPS.

ECLM improves over standard fine-tuning by 8.5% and 21.2% on the respective datasets.

The code for ECLM is publicly available at the provided GitHub link.

Abstract

Large Language Models (LLMs) have demonstrated impressive capabilities in language generation and general task performance. However, their application to spoken language understanding (SLU) remains challenging, particularly for token-level tasks, where the autoregressive nature of LLMs often leads to misalignment issues. They also struggle to capture nuanced interrelations in semantic-level tasks through direct fine-tuning alone. To address these challenges, we propose the Entity-level Language Model (ECLM) framework, which reformulates slot-filling as an entity recognition task and introduces a novel concept, \textit{Chain of Intent}, to enable step-by-step multi-intent recognition. Experimental results show that ECLM significantly outperforms strong baselines such as Uni-MIS, achieving gains of 3.7\% on MixATIS and 3.1\% on MixSNIPS. Compared to standard supervised fine-tuning of LLMs, ECLM further achieves improvements of 8.5\% and 21.2\% on these datasets, respectively. Our code is available at https://github.com/SJY8460/ECLM.