Personality prediction based on social media posts using deep learning algorithms

Abstract

This study analyzes the complicated field of personality prediction through the use of the diverse data embedded in social media posts. Our research uses many different areas of machine learning and deep learning algorithms to study the complicated connection between individual characteristics and digital expressions. The algorithms that were selected include CNN, BiLSTM, LSTM, LR, Linear SVC, DT, RF, and Multinomial Naive Bayes (MNB) deep learning and machine learning architectures. Analyzing these algorithms provides small variations in approach, with every algorithm presenting different points of view on the prediction task. With an accuracy of 81.36%, Linear SVC was the clear winner, closely followed by Logistic Regression at 80.25%. .. The outcomes of this study not only improve the accuracy of personality prediction from social media posts but also determine a basis for future study actions. The combination of deep learning and machine learning to understand the specifics of human behavior on digital platforms has significant promise for a variety of uses, including mental health monitoring and customized advertising methods. The knowledge achieved from this study prepares the way for the responsible and significant application of predictive algorithms in gaining a knowledge of human personalities in the online environment, as technological advances keep changing our digital connections.