Cyberbullying detection with machine learning methods in Turkish texts

Abstract

Undoubtedly, the widespread use of the internet and the increasing popularity of social media platforms have caused the rapid spread of the actions called cyberbullying. The number of people subjected to cyberbullying throughout the world is increasing day by day and it has a great impact on their victims. Identifying this action is of great importance in terms of preventing the emergence of new victims and not being exposed to this action any more. At this point, it has been observed that many studies have been carried out in the literature on the detection of cyberbullying, but it has been determined that the number of studies in Turkish texts is very low. It is thought that this study will contribute to the literature because the dataset used in the study is new and to the best of our knowledge, such a large number of algorithms have not been tested in the literature. In the study, Bagging, Boosting, C4.5, Gradient Boosting, K-Means, KNN, LR, NB, ANN, RO, DVM, Stochastic Gradient Descent and XGBoost algorithms were used comparatively for the first time on this data set. © 2022, Gumushane University. All rights reserved.