ICNLSP 2024: Linking Quran and Hadith Topics in an Ontology using Word Embeddings and Cellfie Plugin

Linking Quran and Hadith Topics in an Ontology using Word Embeddings and Cellfie Plugin
By: Ibtisam Khalaf Alshammari | Eric Atwell | Mohammad Ammar Alsalka
University of Leeds, United Kingdom.
7th International Conference on Natural Language and Speech Processing.
icnlsp.org/202...
Abstract:
Qur'an and Hadith are the sacred texts of the Islamic religion. Arabic Qur'an and Hadith texts have been analyzed and annotated by researchers using a variety of domains, representations, and formats to improve the accessibility of Islamic knowledge. However, the many and diverse Islamic resources raise a potential challenge in linking and integrating them. The main objective of this work is to link Qur'an and Hadith topics and integrate them with related knowledge from different Islamic resources. The proposed methodology is to use a combination of Word embeddings-based BERT with the Cellfie tool to achieve more accurate and meaningful data integration. The results of using the CL-AraBERT word embedding model display efficiency performance in F1 score and accuracy metrics with 91% and 84% respectively. At the same time, the constructed ontology, RQHT, links the Qur'an and Hadith topics with their related knowledge properly and consistently.