Research on Dialogue Generation of Emotion Enhanced Large Language Model Based on Deep Learning

Read full text Summarise with VIRAT
Authors: Xu Wang
Cited by: 0
References: 10

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

As the scenarios of human-computer interaction in China's social development have gradually expanded and giant artificial intelligences (GIIs) have become more frequent, making users feel better through emotional identification and empathic responses has become increasingly significant. Although the current generation of models can achieve precise semantic encoding, they usually face more prominent issues such as emotion-recognition bias and rigid responses; In terms of multi-turn dialogues, there is also a lack of emotional consistency. Due to the above limitations, they are difficult to be adapted for use in emotion-sensitive scenarios such as psychological counselling and intelligent customer service, which affects their application in practice. Based on deep learning technology, this study combined affective computing theory to build an emotion-enhanced dialogue generation system that enhances LLM's ability of empathy and understanding through emotion recognition and adaptive response strategies. Emotion feature extraction, embedding fusion and reinforcement learning optimisation have been used to realise precise alignment of users' real-time emotions by a model. The evaluated deep neural network based on Qwen2.5-7B, daily dialogues and cped datasets has obtained better effects than before, an improvement is expected by increasing emotional adaptability to about 0.86 (from previous level 0.41); Emotion recognition precision increased by more than 25 percent to 0.9 (before) to improve the model performance of users' satisfaction above 79%, making the emotions expressed in the model much closer to those of real people, so this method can be used for improving the warmth and Humanization of Human-machine communication technologies.

Fields of Study & Subjects

Speech and dialogue systems Emotion and Mood Recognition Sentiment Analysis and Opinion Mining

Publication details

Journal / source
Journal of Computer Science and Artificial Intelligence
Volume
6
Issue
3
Pages
42–46
Publication date
29 March 2026
Language
EN