Tu, Rui
(2025)
Large Language Model For Mental Health: Attributional Style Transfer And Data Generation.
Master's Thesis, University of Pittsburgh.
(Unpublished)
Abstract
According to the reformulated version of the Learned Helplessness theory, an individual who experiences uncontrollable negative events may subsequently develop a negative attributional style, whereby the person exhibit greater susceptibility to helplessness deficits in response to negative events compared to their optimistic counterparts. This attributional style not only contributes to depressive symptoms but also represents a malleable target for cognitive therapy. Through attention to patients' attributional style, therapists may be able to alter patients' perceptions and experiences of negative events and thereby decrease the likelihood of depressive symptoms. In an effort to better interpret mental health issues and assist individuals with depressive tendencies in overcoming negative attributions, we introduce the Attributional Style Transfer Dataset (ASTD), which features paragraphs of events described in six distinct attributional styles. Additionally, we offer an open-source benchmark library comprising datasets and baseline methods, designed to support and advance future research and applications in this domain. At the same time, we have experimentally demonstrated that by introducing information from other datasets, rather than simply generating data through LLM, we are able to obtain additional information and improve the diversity of the generated data, which is a breakthrough for the use of LLM for data generation.
Share
| Citation/Export: |
|
| Social Networking: |
|
Details
| Item Type: |
University of Pittsburgh ETD
|
| Status: |
Unpublished |
| Creators/Authors: |
|
| ETD Committee: |
|
| Date: |
7 January 2025 |
| Date Type: |
Publication |
| Defense Date: |
18 November 2024 |
| Approval Date: |
7 January 2025 |
| Submission Date: |
30 October 2024 |
| Access Restriction: |
No restriction; Release the ETD for access worldwide immediately. |
| Number of Pages: |
49 |
| Institution: |
University of Pittsburgh |
| Schools and Programs: |
Swanson School of Engineering > Electrical and Computer Engineering |
| Degree: |
MS - Master of Science |
| Thesis Type: |
Master's Thesis |
| Refereed: |
Yes |
| Uncontrolled Keywords: |
large language model, dataset, mental health, attributional style |
| Date Deposited: |
07 Jan 2025 21:06 |
| Last Modified: |
07 Jan 2025 21:06 |
| URI: |
http://d-scholarship.pitt.edu/id/eprint/47042 |
Metrics
Monthly Views for the past 3 years
Plum Analytics
Actions (login required)
 |
View Item |
![[feed]](/images/feed-icon-32x32.png){kind=icon}