Abstract:
To investigate the acoustic characteristics of Putonghua speech conveying belief and disbelief, and the effects of the Big Five personality dimensions (Neuroticism, Extroverts, Openness, Agreeableness, Conscientiousness) on acoustic distinctions between these two attitudes, the scenarios were designed to elicit utterances in both tones of belief and disbelief from 47 participants. Their personality dimensions were measured using the Neuroticism Extraversion Openness Five-Factor Inventory (NEO-FFI). With six supervised classifiers, speech conveying belief and disbelief was identified on the basis of 23 acoustic features, achieving a maximum recognition accuracy of 0.76. Based on the SHapley Additive exPlanation (Shap) values, eight important acoustic features were selected for subsequent analyses. Redundancy analysis and linear mixed model showed that, compared to belief speech, disbelief speech exhibited higher mean, standard deviation, minimum, and maximum values of fundamental frequency, as well as harmonic-to-noise ratio. In addition, disbelief speech showed lower speech rate, harmonic difference (H1–H2), and Jitter. These acoustic distinctions, however, were moderated by the personality dimensions: Neuroticism affected the mean, minimum and standard deviation of fundamental frequency, while Extroversion influenced the mean fundamental frequency and harmonic-to-noise ratio, and Openness impacted the standard deviation of fundamental frequency. Linear mixed model trees further revealed that the Five Big personality traits significantly modulated the differences in the mean fundamental frequency, the standard deviation of fundamental frequency, and the speech rate between the utterances conveying belief and disbelief.