Reading the Qur'an is an obligation for every Muslim. Before being able to read the Qur'an with correct pronunciation and pronunciation, it is necessary to practice reading letters or hijaiah sentences. One method of learning the Qur'an is using the Iqra method discovered by KH. As'ad Humam. In general, learning activities using the Iqra method require a supervising teacher (ustaz or ustazah) as a student to evaluate the accuracy of the pronunciation of the hijaiah letters. Because the learning process is done face-to-face in the room. So that problems arise, namely the limitations of time, place, and experience. Previous research has tried to overcome this problem, namely by making an Android-based Iqra Scoring Application that can compare the recorded voice of students with the voice of the cleric. However, the results of the overall score similarity correlation between the application assessment and the ustaz's assessment on Iqra 1 is 0.034 in the weak category. To overcome this problem, in this study, we will add the calculation of the timbre score using the Gaussian Mixture Model (GMM) Algorithm and change the score weighting values ​​of each parameter of pitch, volume, rhythm, and timbre to 0.55, 0.05, 0.38, and 0.02. So that the results of the overall score similarity correlation on Iqra 1 is 0.376 in the medium category.
Copyrights © 2021