SISTEM PENILAIAN HAFALAN AL-QUR’AN BERDASARKAN KRITERIA PADA APLIKASI MONITORING SANTRI MENGGUNAKAN FUZZY NEURAL NETWORK

The assessment of Al-Qur'an memorization at Madrasah Diniyah At-Thalibiyah is currently conducted manually, leading to risks of subjectivity, inconsistency, and unstructured documentation. Qualitative assessments using ambiguous terms like "fairly fluent" or "less precise" contain uncertainty that is difficult to measure objectively. This research aims to design and implement an objective and measurable assessment system by integrating a Fuzzy Neural Network (FNN) into a Progressive Web App (PWA) for student monitoring. The research methodology includes data collection through observation and interviews, FNN model design with three input criteria (fluency, tajwid, and makhraj), and application development using the prototype model. The results demonstrate that the developed FNN model is highly effective in automatically classifying memorization predicates, achieving a consistent F1-Score of over 96% across all classes. Furthermore, usability testing of the monitoring application revealed a very high level of acceptance from teachers, with an overall average feasibility score of 95.52%, categorized as "Strongly Agree". This system successfully transforms the manual assessment process into one that is more objective, efficient, and well-documented, while assisting teachers in accurately and transparently monitoring students' memorization progress.
Keywords: Fuzzy Neural-Network, Al-Qur’an Memorizing Assesment, Progressive Web App