Speech-based automatic personality trait prediction analysis

Abstract

Automatic personality perception is the prediction of personality that others attribute to a person in a given situation. The aim of automatic personality perception is to predict the personality of the speaker perceived by the listener from nonverbal behaviour. Extroversion, conscientiousness, agreeableness, neuroticism, and openness are the speaker traits used for personality assessment. In this work, a speaker trait prediction approach for automatic personality assessment has been proposed. This approach is based on modelling the relationship between speech signal and personality traits. The experiments are performed over the SSPNet speaker personality corpus. For speaker trait prediction, support vector machines (SVM), multilayer perceptron (MLP), and instance-based k-nearest neighbour were analysed with multiple features. Various features have been analysed to find suitable feature for various speaker traits. The analyses have been conducted using pitch, formant, and mel frequency cepstral coefficients (MFCC) and the analysis results are presented. The accuracy of 100% has been obtained for MFCC features with 19 coefficients.