<![CDATA[This study focuses on the use of the Qur'an as the primary source of Islamic teachings, aiming to facilitate Muslims' understanding of its content. To achieve this, the classification of translated Qur'anic verses was conducted. Two methods that are rarely used for Qur'anic translation data are Random Forest (RF) and Long Short Term Memory (LSTM) due to their ability to process large and complex data. The data used in this study are translations of the Qur'an that have been classified into 15 topics by previous research, but this study will only focus on 6 topics. The objective of this research is to compare the performance of RF and LSTM in classifying Qur'anic translations into 6 different categories. The results show that in the preaching category, LSTM consistently outperformed RF, with an F1-Score of 57.3% and an accuracy of 96.8%, whereas RF achieved an F1-Score of 49.4% and an accuracy of 97.5%. These findings indicate that LSTM has better performance, especially with proper preprocessing, optimal parameter tuning, and balanced data. This study provides important insights into the development of classification models for Qur'anic translation texts, highlighting the importance of proper preprocessing and parameter tuning.]]>
Copyrights © 2024
