Semantic Abductive Network Construction for the Holy Qur’an: A Hybrid Ontology-Based Approach

The Engineering and constructing semantic networks constitute one of the foundational technologies in the fields of cognitive processing, natural language processing, the semantic web, and the development of artificial intelligence-based systems. Consequently, expertise in the design, construction, engineering, maintenance, evolution, and optimization of ontologies has played a crucial role in advancing intelligent technologies in recent years, and this trend, particularly in the context of dependable and responsible artificial intelligence, is expected to continue in the coming years. The Holy Qur’an, as the sacred book of Muslims worldwide and the primary source of Islamic religion, civilization, and culture, has consistently served as a principal resource in the humanities and Islamic studies, as well as in socio-religious service applications, within Muslim communities. In this paper, a semantic network is automatically constructed using a hybrid approach that integrates multiple technological solutions, including ontologies, word embeddings, co-occurrence analysis, and Arabic root extraction. After the construction of the semantic network and through the application of clustering techniques, several semantic frames were automatically extracted and designated as “abduction frames.” To evaluate the proposed approach, a questionnaire-based assessment was conducted, in which ۱,۲۹۵ individuals participated voluntarily. The results yielded a precision of ۶۹.۴۷% and a recall of ۸۵.۳۵%. Additionally, a mixed quantitative–qualitative evaluation conducted by a panel of experts rated the validity and innovation of the proposed method’s outputs as “good.”

کلیدواژه ها:

the Qur’an ، Ontology ، Word embedding ، Co-occurrence ، semantic frame ، Abduction ، Association ، Collocation