K. Rajagumar, Prishalini (2025) Identifying Suicide Risks Through Twitter (X) In Malaysia. Masters thesis, Perpustakaan Hamzah Sendut.
![[img]](http://eprints.usm.my/style/images/fileicons/application_pdf.png) |
PDF
Download (238kB) |
Abstract
The rise of suicide rates globally highlights the urgent need for improved methods of identifying and understanding suicidal risk, particularly in regions with limited data collection systems. This thesis focuses on the challenges of suicide-related data acquisition in malaysia where stigma and structural limitations such as reliance on medically reported information and the historical criminalization of suicide attempts, which was only decriminalized in 2023 impedes comprehensive reporting. To address this gap, this study explores the potential of twitter as a supplementary data source for detecting suicide expressions through sentiment analysis. The primary objective of this research is to evaluate the potential of utilizing sentiment analysis tool (vader) on platforms such as twitter to detect individuals exhibiting signs of suicide risk through the analysis of tweet content. Utilizing twitter’s api, tweets from malaysian users aged 18 and above, written in english, were collected by filtering for specific keywords. Two keyword lists were used: one targeting suicide expressions and the other focusing on positive statements. The resulting tweets were divided into suicide expression and positive-related (control) datasets and manually coded by a human coder based on content and emotional valence. Sentiment analysis was then applied to both datasets, with its performance benchmarked against human coding to assess accuracy.
| Item Type: | Thesis (Masters) |
|---|---|
| Subjects: | H Social Sciences > H Social Sciences (General) > H1-99 Social sciences (General) |
| Divisions: | Pusat Pengajian Sains Kemasyarakatan (School of Social Sciences) > Thesis |
| Depositing User: | Mr Hasmizar Mansor |
| Date Deposited: | 01 Jul 2026 01:42 |
| Last Modified: | 01 Jul 2026 01:42 |
| URI: | http://eprints.usm.my/id/eprint/64532 |
Actions (login required)
 |
View Item |
