Plantar Pressure Based Personality Type Classification Using Gait Cycle Features

Gait contain diverse information about individuals, and while research on gait and personality has been on the rise recently, there remains a scarcity of comprehensive studies involving various gait data and personality traits. In this study, a total of 24 healthy subjects were recruited, and data on personality types along with plantar pressure during gait were collected. Personality traits used in the study included E-I for interaction with surroundings, S-N for perspective on the world, T-F for how to make decisions and deal with emotions, and J-P for approach to decision-making. After extracting gait cycle features from the collected data, we used a Support Vector Machine (SVM) classification model to classify personality traits and scrutinize feature importance. Furthermore, Pearson correlation analysis based on feature importance was conducted to analyze the correlation between gait cycle features and personality. According to the research findings, classification accuracies for E-I, S-N, T-F, and J-P were 0.775, 0.741, 0.63, and 0.65, respectively. In correlation analysis, E-I exhibited the highest correlation at -0.45 with the 2nd Double Support Time, while S-N showed the highest correlation at -0.418 with Mid Stance Time. However, T-F and J-P did not display significant correlations. This study partially confirms the potential of classifying personality traits solely based on gait cycle features derived from plantar pressure data, providing foundational insights for the analysis of gait and personality.