Personality Dimensions Classification with EEG Analysis using Support Vector Machine

Abstract

Personality is the fundamental thing that forms the behavioral tendencies of each individuality in a situation. A common model used to describe personality is the big five personality that divides personality traits into five dimensions of neuroticism, extraversion, openness, agreeableness, and conscientiousness. Personality assessment through physiological signals offers objectivity and reliability of the test results due to the minimal role of test takers in the examination process. One widely recommended approach is signal-based analysis of electroencephalography (EEG). The EEG signal feature of the ASCERTAIN public database was extracted using discrete wavelet transform (DWT) and was classified using support vector machine (SVM) to determine personality dimensions. The results showed better performance compared to the application of other techniques on the same dataset with 69% and 75.9% accuracy to determine extraversion and neuroticism level, respectively. However, this accuracy still needs to be improved to generate reliable model. Increased data variability can be useful for understanding brain dynamic activity per individual.