| Author: | Li, Junlin |
| Title: | Enhancing empathy with pragmatics through computational modeling of conversation |
| Advisors: | Huang, Chu-ren (LST) |
| Degree: | Ph.D. |
| Year: | 2026 |
| Department: | Department of Language Science and Technology |
| Pages: | xvii, 152 pages : color illustrations |
| Language: | English |
| Abstract: | Empathy is inseparable from dialogic context. Empathetic dialogue systems generate empathetic responses based on dialogic context. From the perspective of linguistics, especially pragmatics, the expression of empathy requires human-aligned knowledge from multiple pragmatic domains, such as micropragmatics (e.g. speech acts), to macropragmatics (e.g. conversation maxims and crosslingual pragmatics). However, the linguistic-pragmatic enhancement of empathetic dialogue systems has been severely ignored in previous studies, leaving an enormous knowledge gap regarding how linguistic-pragmatics may improve the empathetic capacity of automatic conversation systems. This thesis presents four language-engineering studies that incorporate pragmatic-inspired constructs into pre-trained and large language models. Through automatic and human evaluations, we demonstrate profound enhancement of empathetic capability due to the computational modeling of representative pragmatic knowledge, including the expression of speech acts, the "effect-effort" trade-off of relevance, the programming of quantity maxims, and also the universal representation of empathy for crosslingual transfer. The current thesis seals the longstanding research gap with respect to the contribution of linguistic-pragmatic knowledge to empathetic chatbots and human-AI interaction. It offers empirical insights for future studies to extend the application of a wider diversity of linguistic knowledge and theories to enhance the conversational competence of affective dialogue systems and AI chatbots for social good. |
| Rights: | All rights reserved |
| Access: | open access |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/14619