Cyberbullying detection: an ensemble learning approach

Online social networking platforms have become a common choice for people to communicate with friends, relatives, or business partners. This allows sharing life achievement, success, and much more. In parallel, it also invited hidden issues such as web-spamming, cyberbullying, cybercrime, and others. This paper addresses the issue of cyberbullying using an ensemble machine learning model. The complete experiment works in two phases: firstly, k-nearest neighbour, logistic regression and, decision tree classifiers are used to detect the bullying post. Secondly, the prediction outcomes of these classifiers are passed to a voting-based ensemble learning model for the predictions. The experimental outcomes confirmed that the ensemble model is detecting bullying posts with good accuracy.