The development of an automatic emotion recognition technique based on electrophysiological signals while listening to quranic recitation

Abstract

Relaxation and calmness are two emotions that people continually seek. One popular method people frequently used to reduce their tension and pressure levels is listening to various types of relaxing music. However, the Quran is composed of Allah’s words, which were ultimately given for the benefit of humanity. Muslims strongly believe that listening to or reading the Quran brings them comfort, pleasure and confidence that would otherwise elude them; however, scientific evidence is still required to prove that this belief has a scientific basis. Recently, researchers have used electrophysiology to explore the relationships between electrical phenomena and body processes. This research aims to study and analyse the electrical activity of people`s brains and hearts when listening to Quranic recitation compared with listening to relaxing music. Two types of electrophysiology readings are used in this research: electroencephalograms (EEGs) and electrocardiograms (ECGs). An EEG measures brain electrical activity, and an ECG measures heart electrical activity. EEG and ECG data were collected from twenty-five subjects. Then, machine learning algorithms were applied to the EEG and ECG signals. In addition, EEG brainwaves were measured, focusing on the alpha and beta bands. The ECG signal analysis also involved heart rate calculation. All these types of analysis were used to measure subjects’ calmness levels and to recognize their emotions while listening to Quranic recitation as compared with listening to relaxing music. With respect to the valence-arousal analysis result, we conclude that Quranic recitation demonstrated a positive transformation of the subjects` emotions: from negative precursor emotions to calmness and happiness conditions denoted by a positive valence for the EEG and ECG signals. In contrast, relaxing music showed a positive transformation with regard to the valence in the EEG analysis; however, with respect to the ECG music data analysis, the results revealed a negative transformation for most of the music tracks.