Mamdani Method–Based Fuzzy Inference System for Determining the Quality of Qur’an Memorization

Abstract

Assessing the quality of Al-Qur'an memorization is an important part of tahfidz learning that requires objectivity and consistency. In practice, memorization assessment is often subjective because it depends on the examiner's perception and uses linguistic criteria that do not have clear boundaries. This condition necessitates a systematic approach that can accommodate uncertainty in the assessment process. This study aims to apply the Mamdani Fuzzy Inference System method in determining the quality of Qur'an memorization objectively. The system is designed with three input variables, namely memorization fluency, tajwid accuracy, and fashahah, and one output variable representing memorization quality. The Mamdani method is selected due to its intuitive rule structure, which aligns with human reasoning. The inference process consists of fuzzification, rule formation, implication, aggregation, and defuzzification. The system employs three input variables with three fuzzy sets each, resulting in 27 IF–THEN rules. The min operator is used for implication, the max operator for aggregation, and the centroid method for defuzzification. The proposed system was implemented using MATLAB and validated through simulation-based case scenarios representing various combinations of input values. The results show that the Mamdani Fuzzy Inference System is able to comprehensively integrate the three assessment aspects and produce a more consistent evaluation of memorization quality compared to conventional averaging methods. This consistency is reflected in the stability of output values generated by the system, where similar input patterns produce identical outputs (e.g., 75.00 and 91.33), whereas conventional calculations yield more fluctuating results. Therefore, the developed model can serve as an alternative decision support system in evaluating Al-Qur'an memorization learning.