Forecasting traits: human personality prediction with machine learning methodology - a comparative study

Data science and AI are transforming the planet through technological change. We can look at many applications of machine learning in our daily lives, but one of the most widely used applications of machine learning is to categorize individuals based on their personality traits. Everyone on this planet is unique and carries a personality—a special kind of character. The availability of high-quality data has paved the way for more effective marketing campaigns targeted at specific demographic groups for predictive analytics and behavioural psychology. Such personality-based communication is more effective in creating more acceptable and appealing products and services. This research work proposes a framework for predicting personality using big five personality traits. Throughout the globe, people are exposed to competitive exams that focus mainly on their personality evaluation. The main motivation behind these assessments is to test the personality and skill sets of people. From personality classification, one can also determine the type of personality and improve the personality based on the results. In this context, we have utilized machine learning models such as Random Forest Classifier, Decision Tree Classifier, Ada-Boost Classifier and hybrid XGB - eXtreme Gradient Boosting Classifier for forecasting traits – personality prediction. The investigations were validated with four performance metrics; accuracy, precision, recall and f1 score. Throughout the experimentations, the hybrid XGB Classifier outperformed all the other algorithms with an accuracy of 95.66%. The precision, recall and f1-success of the hybrid XGB Classifier were evaluated to be 96% respectively.