Personality traits can be used to explore individuals’ potential needs in different contexts, such as driving. However, relying on traditional method of measuring personality traits (e.g. self-report questionnaires) not only does not meet drivers’ needs for vehicle adaptation but also takes considerable time and concentration in the scenario of nonfixed drivers (e.g. taxis, rental cars, and family cars) .

Recently, a research group led byProf. ZHU Tingshao from Key Laboratory of Behavioral Science, Institute of Psychology of the Chinese Academy of Sciences collected driving signals provided by in-vehicle sensors of 92 participants on a 15 km pre-defined route (Figure 1) and built machine learning models to identify driver’ Big Five personality traits automatically.

Figure 1. Pre-defined route during data collection. Image by WANG Yameng.

The researchers extracted features from raw driving signals in the time and frequency domains respectively using statistical methods and the discrete Fourier transform. These features will be then used to identify Big Five personality traits by machine learning approach.

The results showed that personality traits can be revealed through driving signals, and time-frequency features extracted from driving signals are effective in characterizing and identifying Big Five personality traits. This approach could be of potential value in the development of in-car integration or driver assistance systems and indicates a possible direction for further research on convenient psychometric methods.

This work entitled "Identifying Big Five Personality Traits through Controller Area Network Bus Data" was published in Journal of Advanced Transportation on Oct 2020.

This work was funded by the BMW China Research Projec, National Natural Science Foundation of China, and Youth Innovation Promotion Association CAS.

Contact:
Ms. LIU Chen
Institute of Psychology
Email: liuc@psych.ac.cn