Personality Distinguish Based on Text Messages Using Machine Learning

Abstract

Day after day, the number of users of text messages on social media is growing, knowing that, the impersonation on these sites is also growing, and often the identity of the sender is indistinguishable. In order to get rid of blackmail and threats and protect the users of these sites by recognizing text messages dialogues, this paper is therefore an attempt to identify the individual through personal message dialogues. The text selection method is used and its classification performance is verified using six machine learning methods (Random Forest, Decision Tree(J48), KNN, Logistic Regression, Naive Bayes, and SGD). The results showed that the Decision Tree and Random Forest outperformed other classification methods with a precision of about 99%.