A Comparison Between Twitter Based Naïve Bayes and Artificial Neural Network Comment Classification Algorithms

Abstract

Machine learning has a wide range of uses, and one of its key uses is classification. A new observation is classified to determine which category it belongs to. Classifiers are the common name for machine learning classifiers. A classifier's task is to use training data provided to it to determine the relationship between a given input variable and a specific group that has already been identified by the system. Perceptron, Naive Bayes, Decision Tree, Logistic Regression, K-Nearest Neighbor, Artificial Neural Networks/Deep Learning, and Support Vector Machine are a few of the techniques used for training classifiers. However, every algorithm has unique benefits and drawbacks. As a result, it is necessary for us to determine which technique between naive bayesian classifiers and artificial neural networks Perform best for classifying tweets. Naive Bayes classifiers are Bayes theorem-based classifiers. The Nave Bayes algorithm is based on the assumption that for a training comment, C the classifier computes the probability that the comment should be categorized under Ki, where Ki is the ith category. On the other hand, an artificial neural network is a mathematical model that attempts to mimic the composition and operation of biological neural networks. Separating the problem's classes using just training data is the aim of supervised classification algorithms. The Artificial Neural Network Showed the highest precision of 0.97 in the regular Twitter data while the Naive Bayes model Showed the highest precision (0.85) in the data on hate speech. However, both algorithms have a recall of 0.96 in the weighted average data.