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.
