Detecting Turkish Cyberbullying Tweets Using Machine Learning

Abstract

Cyberbullying is a form of crime where individuals are subjected to online hate speech and harassment, and its prevalence has increased with the growth of social media. There is a noticeable gap in the current literature, especially for cyberbullying detection in languages other than English. This study proposes a method for automatic cyberbullying detection in Turkish tweets. The proposed model incorporates the Support Vector Machine and Random Forest classification algorithms. The model has been trained on labeled real-world data sourced from Twitter. To address the characteristics of the Turkish language, a natural language processing tool called Zemberek-NLP has been used. This tool captures the nuances of the language, enhancing the accuracy of the detection model. This research aims to contribute to the fight against cyberbullying by presenting an innovative approach to detecting it in Turkish.