IMPLEMENTASI AUTOMATIC SPEECH RECOGNITION PADA PENILAIAN BACAAN AL-QURAN DENGAN METODE MUROJA’AH BERBASIS ANDROID

This research focuses on developing an Android application for assessing Al- Qur'an recitation utilizing Automatic Speech Recognition (ASR) technology and the muroja'ah method. The application aims to address the issue of low proficiency in Al-Qur'an recitation by providing users with constructive feedback on their pronunciation. The ASR feature is implemented using Google Speech API and the Jaro-Winkler algorithm to measure the similarity between the user's recitation and the correct Al-Qur'an text, resulting in an average score of 80% for recitations that pay attention to the length of the readings and 65% for those that do not in Surah An-Nas, as well as 86% for recitations that pay attention to the length of the readings and 70% for those that do not in Surah Al-Fil. The muroja'ah feature allows users to set reminders for regular repetition of Al-Qur'an recitation. Black- box testing results indicate that all application features function correctly. Usability testing on 30 respondents revealed a high level of acceptance, with 80% of respondents finding the application easy to learn and efficient, and 83.3% expressing satisfaction with the application. This research demonstrates the potential of ASR-based applications in improving Al-Qur'an recitation skills and providing a more effective and personalized learning experience.

Keywords: Automatic Speech Recognition, Al-qur’an, Android, Rapid Application Development