•  
  •  
 

Document Type

Article

Keywords

Facebook Groups, Cyberbullying, Cybercrime, Text Mining, Logistic Regression (LR)

Abstract

The widespread utilization of social media platforms such as Facebook, Twitter, and Instagram inside academic organizations has become fundamental for student correspondence and joint effort, yet it has simultaneously prompted an increase in cyberbullying incidents. Cyberbullying is embodied by offensive and harmful comments that sabotage the casualty's prosperity. This study plans to resolve this issue by fostering an extensive module to distinguish cases of cyberbullying inside Facebook groups of Mosul University students. Our methodology starts with the collection of data in the Arabic language, which is then exposed to careful manual handling to label comments and eliminate clamor-like copies and superfluous sections. The dataset, comprising 2,715 comments, goes through prehandling and element extraction via the term frequency-inverse document frequency (TF-IDF) strategy. Consequently, we utilize the logistic regression (LR) classifier to investigate and order the data. Our findings revealed that more than 21% of the comments dissected were brutal, a disturbing rate for a local area such as Mosul University. These outcomes highlight the critical requirements for intercession and the significance of observing student cooperation on social media. The dataset and experiences obtained from this study are important for psychologists and psychological direction experts to comprehend and relieve cyberbullying. This exploration not only highlights the prevalence of cyberbullying in academic settings but also provides a strong systemic structure for future examinations to expand, highlighting the basic role of AI in battling web harassment.

Download

Share

COinS