A Response Generation Framework Based on Empathy Factors, Common Sense, and Persona

Building a human-like dialogue system is a challenging task that requires effective use of context, common sense and personal information. In a conversation, the responder usually analyzes the emotion, intention, and common sense involved in the speaker's sentence. Based on this analysis, the r...

Full description

Saved in:
Bibliographic Details
Published in:IEEE access 2024, Vol.12, p.26819-26829
Main Authors: Li, Weijie, Yang, Yong, Tuerxun, Palidan, Fan, Xiaochao, Diao, Yufeng
Format: Article
Language:English
Subjects:
Comets
Communication systems
Context modeling
Data models
Emotion recognition
emotional responses
Emotions
Empathy
Frequency response
History
IP networks
Natural language processing
Oral communication
Personal information
System effectiveness
Vocabulary
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Building a human-like dialogue system is a challenging task that requires effective use of context, common sense and personal information. In a conversation, the responder usually analyzes the emotion, intention, and common sense involved in the speaker's sentence. Based on this analysis, the responder considers both the above-mentioned content and their personal information to formulate a response. Previous work in this area has only focused on one or some aspects, such as emotion, intention, common sense or persona, rather than considering all of them together. To address this issue, we propose a response generation framework called EFCP, which is based on empathy factors, common sense, and persona. This framework simulates a rich dialogue generation process that is rarely seen in previous work. In predicting the type of empathy factors a responder should adopt, we consider both the responder's personal information and the conversation history. Our experiments show that this method effectively improves the accuracy of prediction. EFCP outperforms the baseline on a variety of automatic metrics and manual metrics, showing its potential for building more effective and human-like dialogue systems.
ISSN:2169-3536
2169-3536
DOI:10.1109/ACCESS.2024.3365533