Machine Learning-Based Detection of Cyberbullying in Social Media Texts: A New Zealand Contextual Study
Al Hassani, Oras, Subaramaniam, Kasthuri, Palaniapan, Sellappan, Lopez, Dobrila and Bin Jalil, Abdurrahman (2026) Machine Learning-Based Detection of Cyberbullying in Social Media Texts: A New Zealand Contextual Study. ICAROB2026, 31. ISSN 2188-7829
Abstract
This study investigates the use of machine learning approaches to identify cyberbullying within social media text, focusing on New Zealand. The research develops and evaluates supervised machine learning models trained on a labelled dataset of X (Twitter) posts obtained from Kaggle. Text preprocessing involved data cleaning, tokenisation, lemmatisation using WordNetLemmatizer, and vectorisation through Term Frequency–Inverse Document Frequency (TF–IDF). Several algorithms, including Support Vector Machine (SVM), Naïve Bayes, Logistic Regression, and Random Forest, were implemented and compared using performance metrics. A Streamlit-based web application was developed to operationalise the most effective model, allowing users to detect cyberbullying in real time.
Actions (login required)
 |
Edit Item |