Intelligent Web Service System for Detecting Cyberbullying on Twitter Based on Support Vector Machine and Random Forest Algorithms

Cyberbullying is a prevalent form of digital crime within the social media sphere. A survey conducted in 2021 among 6,000 individuals aged 10 to 18 showed that 50% of the surveyed children had encountered instances of cyberbullying. Approximately 47% of participants recounted instances where they faced threats, derogatory comments, and offensive messages on social media platforms. These acts of intimidation and disparagement can significantly harm mental well-being. The integration of machine learning methodologies, such as Support Vector Machine (SVM) and Random Forest, was used as a viable approach to predict whether a given social media post, particularly tweets, falls under cyberbullying. This research starts with data collection, followed by tokenization, modeling, evaluation and deployment of web service. The research attained 84.24% and 84.98% accuracy using Random Forest and Support Vector Machines, respectively. Both algorithms produced models that were subsequently deployed as a web service API through various multiplatform applications. The findings of this study are available for reference at uns.id/winarSVMRF.