One application of the voice recognition technology is as an automatic answering machine to help a receptionist. In this research, a gender identification system has been created by utilizing speech signals to help improve interaction between automatic answering machines and customers. The identification process is carried out by comparing the incoming voice features with the speech signal feature data set that has been stored as a code book. The feature extraction used in this study is the Mel-Frequency Cepstral Coefficients (MFCC) because this method is a feature extraction method that approaches the human hearing system. Furthermore, the results of this feature extraction will be used to classify male or female gender using the Decision Tree algorithm, so that the gender of the speaker can be determined. The accuracy of gender prediction results for feature extraction data for vocal a data reaches 100%, vocal i data reaches 96.88%, vocal u data reaches 81.25%, vocal e data reaches 90.63%, and vocal o data reaches 87.50%.
Copyrights © 2023