Enhancing Named Entity Recognition in Modern Standard Arabic via Fine-Grained Part-of-Speech Tags
Freihat, Abed Alhakim Abbas, Mourad Alfraidi, Tareq Alluhaibi, Reyadh Sultan Al-Thubaity, Abdulmohsen
Freihat, Abed Alhakim
Abbas, Mourad
Alfraidi, Tareq
Alluhaibi, Reyadh Sultan
Al-Thubaity, Abdulmohsen
Supervisor
Department
Natural Language Processing
Embargo End Date
Type
Book chapter
Date
2025
License
Language
English
Collections
Research Projects
Organizational Units
Journal Issue
Abstract
Arabic, a morphologically rich language, poses unique challenges for named entity recognition (NER) due to its lack of capitalization, complex word forms, and significant ambiguity. This study investigates the impact of incorporating fine-grained POS tags in ANER, demonstrating an F1 score improvement from 74.2% to 87.5% across four datasets with increasing annotation complexity. These findings highlight the importance of linguistic context in improving ANER and suggest broader applications in machine translation, sentiment analysis, and other NLP tasks.
Citation
A. Alhakim Freihat, M. Abbas, T. Alfraidi, R. Sultan Alluhaibi, and A. Al-Thubaity, “Enhancing Named Entity Recognition in Modern Standard Arabic via Fine-Grained Part-of-Speech Tags,” Signals and Communication Technology, vol. Part F1026, pp. 119–139, 2025, doi: 10.1007/978-3-031-93612-8_6
Source
Signals and Communication Technology
Conference
Keywords
Arabic, Arabic named entity recognition, Machine learning, Named entity recognition, Sequence labeling
Subjects
Source
Publisher
Springer Nature