Studying and Comparing the Effectiveness of Audio Features Extraction Algorithms for Recognize Openness Personality

Audio signal includes many information about speaker as social background as dialect, gender, age, and mental state of short and long term mental state, which represents the focus of research. Research aims to design a system to highlight the speaker's personal style, depending on attributes governing long term psychiatric included in audio signal, consisting of a database containing the audio files to a variety of speakers in terms of personal style according to BFI test – 10 and five factor theory, where the audio database Completed recordings included going back to 75 male and 75 female university students, all of the convergent reconstruction and do not suffer from flaws or receive audio. The research analyzed the voice features speakers recordings using feature extraction algorithms, LPC, MFCC, LSP and found that people loose construction lists be MFCC transaction values in the vehemence of high voice, while the LPC parameters, the LSP did not show distinct signs govern any personality style.