Graduate Thesis Or Dissertation


Scalable and Timely Detection of Cyberbullying in Online Social Networks Public Deposited

Downloadable Content

Download PDF
  • The exponential growth of popularity of online social networks in the last decade has unfortunately paved the way for the threat of cyberbullying to rise to an unprecedented level. So a research that provides insights into the analysis of cyberbullying incidents and building a system that is highly scalable and responsive is of unparalleled need. This dissertation gathers insights into cyberbullying incidents in video and image-based social networks (Vine and Instagram respectively) and then presents a system solution that makes use of the gained insights to improve efficiency and efficacy of cyberbullying detection. First, it presents detailed analyses of cyberbullying incidents in the Vine social network by collecting data and labeling them by CrowdFlower. Second, it performs a thorough investigation of the differentiating factors of cyberbullying in online social networks. Third, it implements a highly scalable and responsive system solution for cyberbullying detection along with a comprehensive evaluation of its performances in terms of timeliness and scalability against a highly popular online social network. Fourth, it outlines design, implementation and preliminary user experience analysis of an android application, BullyAlert, that was developed to enable guardians to get adaptive notifications for cyberbullying based on their individual subjective tolerance levels. Finally, it shows that using textual and video feature greatly improves cyberbullying detection classifier's performances.
Date Issued
  • 2018
Academic Affiliation
Committee Member
Degree Grantor
Commencement Year
Last Modified
  • 2019-11-14
Resource Type
Rights Statement