Representation of dataset frequency based on Text column

Representation of training and testing data ratio

Diagram of proposed Methodology

Result:

Model	Accuracy	Precision	Recall	F1-score
Logistic Regression	72.02%	0.70	0.66	0.68
Multinomial NB	70.89%	0.68	0.66	0.67
Random Forest	70.55%	0.67	0.66	0.67
Decision Tree	65.00%	0.61	0.62	0.61
K-Nearest Neighbour	61.60%	0.57	0.58	0.58

Representation of ROC-AOC

Result Anlysis :

Based on the results, Logistic Regression and Multinomial Naive Bayes have higher accuracy, precision, recall and F1-score compared to Random Forest, Decision Tree and K Neighbors classifiers for detecting cyber bullying using Bangla corpus.