Speech Emotion Recognition Model using Support Vector Machine Through MFCC Audio Feature

Emotions are one of the most influential aspects of everyday life. Everyone has their own way of expressing their emotions. One way to express emotions is through speech or by speaking. This gave rise to a new field of research called speech emotion recognition which aims to understand a person’s emotions through sound. In this study, the Support Vector Machine algorithm will be implemented to create a speech emotion recognition model through the MFCC voice feature obtained from voice processing. The model created can be used to classify six emotions, namely happy, sad, angry, fear, disgust, and neutral. The highest accuracy is obtained from the model created using the Support Vector Machine algorithm using a radial basis function kernel which is considered to be able to properly classify emotions based on sound. The usage of the combined dataset also improved the accuracy of the model and are able to obtain above 70% highest accuracy on each test.