Cyberbullying Detection using SVM Algorithm
International Journal for Research in Engineering Application & Management · 2024
Cyberbullying has emerged as a significant societal issue prevalent on the internet, impacting both adolescents and adults alike. Its detrimental effects include instances of suicide and depression among victims. Consequently, there is an escalating necessity for regulating content across social media platforms. This study endeavors to address cyberbullying through the utilization of Natural Language Processing (NLP) and Machine Learning (ML) techniques, employing data sourced from two distinct forms of cyberbullying: hate speech tweets from Twitter and personal attack comments from Wikipedia forums. The research aims to construct a model for detecting cyberbullying in textual data. Specifically, three distinct methods for feature extraction and four classifiers are scrutinized to ascertain the optimal approach. The evaluation of the model indicates that for tweet data, accuracies surpassing 90% are achieved, while for Wikipedia data, accuracies exceed 80%. This research contributes to the ongoing efforts in combating cyberbullying through advanced computational techniques.