Real Time Cyberbullying Detection
Shalni Prashar1, Suman Bhakar2

1Shalni Prashar,  Pursuing Post-Graduation Student, Computer Science program, Rajasthan College of Engineering for Women, Jaipur, India.
2Suman Bhakar,  Assistant Professor, Computer Science Department, Rajasthan College of Engineering for Women, Jaipur, India.
Manuscript received on November 26, 2019. | Revised Manuscript received on December 15, 2019. | Manuscript published on December 30, 2019. | PP: 5197-5201  | Volume-9 Issue-2, December, 2019. | Retrieval Number: B4253129219/2019©BEIESP | DOI: 10.35940/ijeat.B4253.129219
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (

Abstract: Automated approaches for detecting cyberbullying on online platforms has remained a primary research concern over past years. Cyber bullying is defined as the use of electronic communication to bully a person, typically by sending messages of intimidating or threatening nature. The victims especially teenagers suffer from loss of confidence, depression, sleep disorder. The research on automated cyberbullying approach is mainly focused on data driven methods. Such methods work on a database of static texts, usually collected from online platforms and are not feasible for dynamic nature of a real-life social networking scenarios. The aim of our research is to develop a cyberbullying detection system using Fuzzy Logic. Three types of bullying emotions are considered in this research work namely aggression, abuse and threat. In the proposed approach chat between two users is continuously monitored and emotion present in each message is determined. Based on the emotion each user’s behavior is categorized as decent or bullying. If the detected bullying nature is higher than a defined threshold value the account of user is ceased and reported automatically. The proposed approach is tested with a chat application developed in Microsoft .Net Framework and approach can detect cyber bullying in good time. The proposed approach, if implemented with social networking platforms can serve as a useful aid for preventing online harassment. The developed algorithm can also be applied in surveillance and human behavioral analysis.
Keywords: Cyber bulling, Fuzzy Logic, k means clustering, Abusive language detection.