Date of Award

Spring 5-16-2026

Degree Type

Thesis

Degree Name

MS Data Science

Department

Mathematics and Computer Science

Advisor

Shajina Anand, Ph.D.

Committee Member

Manfred Minimair, Ph.D.

Committee Member

Nathan Kahl, Ph.D.

Keywords

Cyberbullying, Sentiment Analysis, Ontology, Natural language processing, BERT, Mental health

Abstract

Cyberbullying has risen as a worldwide concern in higher education because of the widespread use of social media and digital communication platforms among college students. Studies have shown that exposure to online harassment can negatively affect students’ mental health, academic performance, and overall well-being. Despite the increasing attention to cyberbullying in recent years, its impact on college students on emotional outcomes and recovery experiences remains underexplored. This study investigates how cyberbullying affects the mental health of college students using multiple approaches that integrate survey data with text analysis. Data was collected through anonymous surveys and interviews among undergraduate and graduate students at Seton Hall University. Data was cleaned and analyzed to uncover relationships between cyberbullying exposure, platform usage, reporting behavior, and self-reported mental health outcomes. Natural Language Processing (NLP) was applied to students’ written descriptions of cyberbullying incidents. Sentiment analysis using a pretrained BERT-based language model was implemented to measure the emotional tone and polarity of cyberbullying messages, and an ontology was developed to categorize cyberbullying behaviors such as insults, threats, harassment, exclusion, and public shaming. The results of this research reveal that cyberbullying among college students is most experienced through popular public social media platforms and is often associated with negative emotional responses. By combining survey and interview analysis with sentiment and ontology text analysis, this research gives a look into cyberbullying experiences of students at Seton Hall and highlights the importance of improved support and prevention strategies for affected students.

Available for download on Tuesday, December 01, 2026

Included in

Data Science Commons

Share

COinS