Supporting Application Fast Learning of Kitab Kuning for Santri' Ula Using Natural Language Processing Methods

Abstract

Education in Islamic boarding schools is one of Indonesia's traditional forms of education that teaches Islamic religious teachings, including studying the yellow classic books as the primary source of spiritual learning. However, learning the Yellow classic book is often complicated by 'ula students (early level students) because Arabic is without harakat or lines, and the material studied is very complex. To overcome these challenges, this research aims to develop a yellow Islamic classic book learning support application for 'ula students using the Natural Language Processing (NLP) method. This application has an interactive chatbot feature that helps students understand the contents of the yellow book more effectively and enjoyably. The research method includes literature study, data collection, data processing, and system development using the Sparse Categorical Cross Entropy algorithm in Natural Language Processing to improve the accuracy of chatbot responses. This application provides an innovative solution by presenting an interactive learning experience that can be accessed anytime and anywhere, thus facilitating Santri learning outside the boarding school environment. The results show that learning for 'ula students with the Natural Language Processing method is very good and easy to understand. The test shows that the accuracy of the application reaches 100% with a low error value (loss), which is 0%. It can be recognized that the effectiveness of Natural Language Processing in supporting yellow book learning, maintaining the tradition of Islamic education in the digital era, and helping teachers and parents monitor the development of students.