Multi-label classification of Indonesian qur'an translation using long short-term memory model

Akbar, Ismail, Faisal, Muhammad ORCID: https://orcid.org/0000-0003-4884-7254 and Chamidy, Totok (2024) Multi-label classification of Indonesian qur'an translation using long short-term memory model. Kinetik : Game Technology, Information System, Computer Network, Computing, Electronics, and Control, 9 (2). pp. 119-128. ISSN 2503-2267

Text 1901-Article Text-124128308-1-10-20240531.pdf Download (539kB)

Abstract

Studying the Quran is an integral act of worship in Islam, necessitating a nuanced comprehension of its verses to ease learning and referencing. Recognizing the diverse thematic elements within each verse, this research pioneers in applying Deep Learning techniques, specifically Long Short-Term Memory (LSTM) and Bidirectional LSTM (Bi-LSTM), coupled with Word Embedding methods like Word2Vec and FastText, to refine the multi-label classification of the Quran's translations into Indonesian. Targeting core thematic categories such as Tawheed, Worship, Akhlaq, and History, the study aims to elevate classification accuracy, thereby enhancing the textual understanding and educational utility of the Quran's teachings. The employment of Bi-LSTM in conjunction with FastText and meticulous hyperparameter optimization has yielded promising results, achieving an accuracy of 71.63%, precision of 64.06%, recall of 63.60%, and a hamming loss of 36.17%. These outcomes represent a significant advancement in the computational analysis of religious texts, offering novel insights into the complex domain of Quranic studies. Furthermore, the research accentuates the critical role of selecting suitable word embedding techniques and the necessity of precise parameter adjustments to amplify model performance, thereby contributing to the broader field of religious text analysis and understanding. Through such computational approaches, this study not only fosters a deeper appreciation of the Quran's multifaceted teachings but also sets a new precedent for the interdisciplinary integration of Islamic studies and artificial intelligence.

Item Type:	Journal Article
Keywords:	LSTM; Bi-LSTM; Word Embedding; Qur’an; Translation Classification
Subjects:	08 INFORMATION AND COMPUTING SCIENCES > 0806 Information Systems > 080607 Information Engineering and Theory
Divisions:	Graduate Schools > Magister Programme > Graduate School of Informatics Engineering
Depositing User:	Muhammad Faisal
Date Deposited:	28 Jun 2024 09:56

Downloads

Origin of downloads

Actions (login required)

View Item