Detecting Cyberbullying in Text Using TF-IDF Vectorization and Support Vector Machine Classification

Cyberbullying is a prevalent form of harassment in India. It encompasses various behaviors that degrade human body features and sentiments in every conceivable way. Victims of cyberbullying often experience psychological distress such as depression and agitation. In response to numerous cases of suicide and public outcry against cyberbullying, the Indian government has implemented punitive measures. Individuals found engaging in cyberbullying can now face legal consequences. To address this issue, we have developed a website capable of detecting cyberbullying based on factors such as generation, cultural background, gender identity, and faith. The ML model we have used here is SVM, which classifies texts and makes predictions on test data with the help of TF-IDF. TF-IDF converts text into numerical values to facilitate fitting the data into the training model. Additionally, in the preprocessing of the data, NLTK is used to convert the text into base forms to classify the data into the aforementioned categories.