Dynamics of Cyberbullying on Twitter: ML Detection Models and the Catalytic role of Tweets Engagement Metrics

In today's interconnected digital landscape, the alarming rise of cyberbullying has cast a shadow over online interactions, significantly impacting the well-being and safety of users. This research conducts a comprehensive analysis of various techniques and algorithms for identifying cyberbullying in Twitter tweets, utilising a diverse dataset and evaluation parameters such as precision, recall, Fl-score, and accuracy. The study assesses the effectiveness of two key detection algorithms, Support Vector Machines (SVM) and Naive Bayes classifiers, leveraging unigram, bigram, and trigram features. Additionally, the research explores how social engagement metrics like likes, retweets, and shares may amplify cyberbullying behaviour, shedding light on the intricate dynamics of cyberbullying detection and its amplification on Twitter through these engagement metrics