Automatic term extraction for Arabic text: Approaches, techniques, and challenges
Muhammed, M., Azab, S., Ali, N., Gheith, M. (2025). Automatic term extraction for Arabic text: Approaches, techniques, and challenges. EKB Journal Management System, 2(2), 51-64. doi: 10.21608/jaiep.2025.423298.1025
Mariam Muhammed; Shahira Azab; Nesrine Ali; Mervat Gheith. "Automatic term extraction for Arabic text: Approaches, techniques, and challenges". EKB Journal Management System, 2, 2, 2025, 51-64. doi: 10.21608/jaiep.2025.423298.1025
Muhammed, M., Azab, S., Ali, N., Gheith, M. (2025). 'Automatic term extraction for Arabic text: Approaches, techniques, and challenges', EKB Journal Management System, 2(2), pp. 51-64. doi: 10.21608/jaiep.2025.423298.1025
Muhammed, M., Azab, S., Ali, N., Gheith, M. Automatic term extraction for Arabic text: Approaches, techniques, and challenges. EKB Journal Management System, 2025; 2(2): 51-64. doi: 10.21608/jaiep.2025.423298.1025

Automatic term extraction for Arabic text: Approaches, techniques, and challenges

Journal of Artificial Intelligence in Engineering Practice
Volume 2, Issue 2, November 2025, Pages 51-64    PDF (752.57 K)
Document Type: Original Article
DOI: 10.21608/jaiep.2025.423298.1025
Authors
Mariam Muhammed* 1; Shahira Azab2; Nesrine Ali1; Mervat Gheith3
1Department of Information Systems and Technology, Faculty of Graduate Studies for Statistical Research (FGSSR), Cairo University, Giza 12613, Egypt
2Department of Computer Science, Faculty of Graduate Studies for Statistical Research (FGSSR), Cairo University, Giza 12613, Egypt.
3Department of Computer Science, Faculty of Graduate Studies for Statistical Research (FGSSR), Cairo University, Giza 12613, Egypt
Abstract
Automatic Term Extraction (ATE) is an essential task in Natural Language Processing (NLP) that aims to identify domain-specific terms from large corpora. In the context of Arabic, ATE plays an essential role in applications such as ontology construction, dictionary development, information retrieval, and text mining. However, the rich morphological structure, and orthographic ambiguities of Arabic present unique challenges in the process of ATE. This paper provides a comprehensive survey of ATE for Arabic text, with a focus on approaches, techniques, and evaluation strategies. We review rule-based, statistical, machine learning, deep learning, and hybrid methods, reviewing their strengths, limitations, and applicability to Arabic’s linguistic characteristics. We also review challenges that affect the ATE process such as morphological richness, multiword expression extraction, named entity recognition, and the scarcity of annotated corpora. Furthermore, we outline evaluation metrics that are essential for assessing performance in Arabic. This paper aims to support the development of more accurate, adaptable, and domain specific ATE systems for Arabic texts.
Keywords
Arabic Language; Automatic Term Extraction; Arabic Natural Language Processing; Multiword Expressions; Information Retrieval
Statistics
Article View: 8
PDF Download: 2