Personality is one of the fundamental aspects that distinguishes individual behavior, thought patterns, and interaction styles. The extraversion dimension, which is part of the Big Five Personality Traits framework, reflects an individual’s tendency to engage in social interactions with two main poles, namely introvert and extrovert. Identifying personality based on this dimension has various applications, ranging from education to employee recruitment. This study aims to develop a personality classification model based on the extraversion dimension using the Extremely Randomized Trees (ERT) algorithm and to compare its performance with other algorithms, namely Decision Tree, K-Nearest Neighbor (KNN), and Support Vector Machine (SVM). The dataset used in this study was obtained from the Kaggle platform, consisting of 2,900 entries and including social behavior indicators represented by five numerical variables and two categorical variables. The research methodology involves data preprocessing, exploratory data analysis, model construction, and evaluation using confusion matrix, precision, recall, F1-score, accuracy, and ROC-AUC. The results indicate that ERT achieves the best performance compared to the other algorithms. The ERT model obtained an accuracy of 92.69%, an F1-score of 0.9269, and a ROC-AUC of 0.9429, outperforming SVM (F1 0.9173; AUC 0.9300), KNN (F1 0.9086; AUC 0.9146), and Decision Tree (F1 0.8879; AUC 0.8876). The superiority of ERT is supported by its tree-based ensemble mechanism with high randomization, which enhances generalization, reduces variance, and captures complex non-linear interactions among behavioral variables. Therefore, ERT is proven to be effective in consistently distinguishing introvert and extrovert tendencies.