AUTOMATIC DETECTION OF CYBERBULLYING IN SOCIAL NETWORKS

Abstract

The increased use of the internet in modern times has created a lot of data. The virtual world has both positive and negative aspects. One of the worst sides of web 4.0 is cyberbullying, a form of cybercrime. The use of technology in bullying situations is known as cyberbullying. This research compiled the findings of thirty separate investigations into cyberbullying and documented the wide range of methods used to identify instances of bullying. By adding textual, behavioral, and demographic aspects, this research departs from a previous analysis of the same dataset that only included linguistic variables. Some terms in the text could be used in a way that could be considered cyberbullying. People who have experienced bullying firsthand are more prone to bullying others. A person's gender, age, and residence are some of the demographic details provided by the dataset. In order to evaluate the two classifiers, the system makes use of various performance metrics. With an overall accuracy of 87.14, the Support Vector Machine classifier outperforms the Bernoulli NB classifier.