Improving the Accuracy by Comparing the Gaussian Naive Bayes Algorithm and Logistic Regression for Predicting Hate Speech Recognition

Comparing the Novel Gaussian Naive Bayes method with Logistic Regression for Hate Speech Prediction with the aim of increasing accuracy. Materials and Methods: This proposed Machine learning uses improved accuracy through Novel Gaussian Naive Bayes algorithm compared with Logistic Regression. This study mainly consists of two groups. There are a total of 20 samples with each set having a example size of 10. The example size calculation is done with a G-power pretest of 80 percent, threshold 0.05% and CI - 95%. Results: When compared to the present system, which has an accuracy of only 92%, it was found that the proposed method obtained 95 % accuracy. There exists no significance between the Novel Gaussian Naive Bayes algorithm Model and the Logistic Regression Model grounded on a sample T-test with the two tailed significance value of $p=0.000$ ($p\lt0.05$). This suggests that the two algorithms vary in a way that is statistically significant. Conclusion: Improved prediction accuracy of hate speech is shown by the outcomes of the Logistic Regression method. Hate Speech can be predicted by the existing system but using the Logistic Regression (LR) algorithm shows better accuracy to predict Hate Speech. So, the LR algorithm is used to forecast Hate Speech.