University of Twente Student Theses
Understanding Suicidal Tweets Using Sentiment Analysis and Topic Modelling
Manu Chandran Nair, Reshma (2024) Understanding Suicidal Tweets Using Sentiment Analysis and Topic Modelling.
PDF
835kB |
Abstract: | This study examines the complex emotional expressions related to suicidal ideation shared on social media, particularly on X. Using two publicly available datasets, the study analyzes 25,482 tweets classified under "depressed" and "significant disruption" categories. Topic modeling in R Studio identified three main themes: (1) Suicidal ideation and anxiety, (2) Life reflection and negative experiences, and (3) Depression and hopelessness. Sentiment analysis using VADER and SentiArt revealed that most tweets expressed negative emotions such as fear, anger, sadness, and disgust, with minimal positive sentiment. Correlation analysis showed weak linear relationships between themes and sentiments, suggesting subtle tendencies toward negative emotions, particularly in life reflection tweets, which generally exhibited neutral sentiments. These findings underscore the role of social media in providing insights into mental well-being and the need for improved algorithms to detect psychological distress. The study demonstrates the potential of sentiment analysis in the early detection and intervention of suicidal ideation and calls for further research on real-time monitoring to enhance mental health support on platforms like X. |
Item Type: | Essay (Master) |
Faculty: | BMS: Behavioural, Management and Social Sciences |
Subject: | 77 psychology |
Programme: | Psychology MSc (66604) |
Link to this item: | https://purl.utwente.nl/essays/102563 |
Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Repository Staff Only: item control page