How to Retrieve Examples in In-context Learning to Improve Conversational Emotion Recognition using Large Language Models?
This work addresses the problem of low accuracy in subjective emotion recognition tasks for conversational AI applications, representing an incremental improvement in example retrieval methods.
The study tackled the challenge of improving conversational emotion recognition (CER) accuracy using large language models by exploring strategies for retrieving high-quality examples in in-context learning. Results showed that augmented example retrieval outperformed other techniques across IEMOCAP, MELD, and EmoryNLP datasets, highlighting the importance of coherent targeted examples and paraphrasing.
Large language models (LLMs) have enabled a wide variety of real-world applications in various domains. However, creating a high-performing application with high accuracy remains challenging, particularly for subjective tasks like emotion recognition. Inspired by the SLT 2024 GenSER Challenge, this study investigates approaches to improving conversational emotion recognition (CER) by LLMs. Specifically, we explore how to retrieve high-quality examples in in-context learning (ICL) to enhance CER. We propose various strategies based on random and augmented example retrieval and also analyze the impact of conversational context on CER accuracy. Experiments were conducted on the three datasets including IEMOCAP, MELD and EmoryNLP. The results show that augmented example retrieval consistently outperforms other techniques under investigation across all datasets, highlighting the importance of retrieving coherent targeted examples and enhancing them through paraphrasing.