COMPARATIVE EVALUATION OF MACHINE LEARNING CLASSIFIERS FOR CYBERBULLYING DETECTION

Abstract:- - In this paper evaluated the performance of several machine learning classifiers in detecting cyberbullying. The models tested include BaggingClassifier, SGDClassifier, LogisticRegression, DecisionTreeClassifier, RandomForestClassifier, LinearSVC, AdaBoostClassifier, MultinomialNB, and KNeighborsClassifier. The evaluation metrics considered were Accuracy, Precision, Recall, and F1 Score. Among the tested models, the BaggingClassifier emerged as the top performer with an accuracy of 0.928, a precision of 0.964, a recall of 0.925, and an F1 score of 0.944, indicating its high effectiveness and balance between precision and recall. The SGDClassifier followed closely, achieving an accuracy of 0.927, a precision of 0.958, a recall of 0.930, and an F1 score of 0.944, demonstrating excellent performance as well. The LogisticRegression model also showed strong results with an accuracy of 0.926, a precision of 0.964, a recall of 0.922, and an F1 score of 0.943. DecisionTreeClassifier and RandomForestClassifier achieved slightly lower accuracies of 0.923 and 0.919, respectively, but maintained strong precision and recall. LinearSVC had an accuracy of 0.917, while AdaBoostClassifier, MultinomialNB, and KNeighborsClassifier showed lower accuracies of 0.908, 0.893, and 0.858, respectively, indicating their relative ineffectiveness for this task. The results suggest that BaggingClassifier and SGDClassifier are highly reliable choices for cyberbullying detection, with LogisticRegression also being a strong contender

---