Classification of Virtual Harassment on Social Networks Using Ensemble Learning Techniques

Background: Internet social media platforms have become quite popular, enabling a wide range of online users to stay in touch with their friends and relatives wherever they are at any time. This has led significant increase virtual crime from the inception these present day. Users harassed when confidential information about them is stolen, or another user posts insulting offensive comments them. posed threat users, both mentally psychologically. Methods: research compares traditional classifiers ensemble learning classifying harassment networks by using models four different datasets: seven machine algorithms (Nave Bayes NB, Decision Tree DT, K Nearest Neighbor KNN, Logistics Regression LR, Neural Network NN, Quadratic Discriminant Analysis QDA, Support Vector Machine SVM) (Ada Boosting, Gradient Random Forest, Max Voting). Finally, we compared our results twelve evaluation metrics, namely: Accuracy, Precision, Recall, F1-measure, Specificity, Matthew’s Correlation Coefficient (MCC), Cohen’s Kappa KAPPA, Area Under Curve (AUC), False Discovery Rate (FDR), Negative (FNR), Positive (FPR), Predictive Value (NPV) were used show validity algorithms. Results: At end experiments, For Dataset 1, had highest accuracy 0.6923 for algorithms, while Voting Ensemble 0.7047. dataset 2, K-Nearest Neighbor, Machine, all same 0.8769 algorithm, Forest Boosting 0.8779. 3, 0.9243 0.9258. 4, 0.8383, voting obtained an 0.8280. A bar chart was represent results, showing minimum, maximum, quartile ranges. Conclusions: Undoubtedly, this technique assisted no small measure comparing selected as well detecting exposing various forms cyber cyberspace. best weakest revealed.