Cyberbullying Detection Neural Networks using Sentiment Analysis

The advances in technical evolution have given rise to a serious problem of cyberbullying. Cyberbullying is the use of electronic communication to bully a person, typically by sending messages of an intimidating or threatening nature. Social networking sites in particular Twitter is becoming a platform for this type of bullying. Machine learning (ML) techniques have been widely used to detect cyberbullying through detecting some language patterns that are exploited by bullies to attack their victims. Sentiment Analysis (SA) of text can also contribute useful features in detecting offensive or abusive content. Deep learning specifically the Convolutional Neural Networks (CNN) has been used to improve the performance of feature extraction during the detection of cyberbullying process. In this research, a SA model is proposed for recognizing cyberbullying tweets in Twitter web-based media. Convolutional Neural Network, Support Vector Machines (SVM) and Naïve Bayes (NB) are utilized in this model as supervised ML classifiers. The aftereffects of the analyses led on this model demonstrated empowering results when a higher n-grams language models are applied on such tweets in comparison with comparable past exploration. Moreover, the results showed that CNN classifiers have outperformed NB and SVM classifiers in several measures.