MBTI Personality Prediction Using Machine Learning and SMOTE for Balancing Data Based on Statement Sentences

The rise of social media as a platform for self-expression and self-understanding has led to increased interest in using the Myers–Briggs Type Indicator (MBTI) explore human personalities. Despite this, there needs be more research on how other word-embedding techniques, machine learning algorithms, imbalanced data-handling techniques can improve results MBTI personality-type predictions. Our aimed investigate efficacy these by utilizing Word2Vec model obtain vector representation words corpus data. We implemented several approaches, including logistic regression, linear support classification, stochastic gradient descent, random forest, extreme boosting classifier, cat classifier. In addition, we used synthetic minority oversampling technique (SMOTE) address issue showed that our approach could achieve relatively high F1 score (between 0.7383 0.8282), depending chosen predicting classifying personality. Furthermore, found SMOTE selected models’ performance (F1 between 0.7553 0.8337), proving integrated with predict classify personality well, thus enhancing understanding MBTI.