Islamic Ontology Coverage Evaluation

Abstract

Ontology has become a popular knowledge representation approach extensively utilized in various computational models. Recently, Islamic research has attempted to employ ontology to capture the massive amount of knowledge in the Quran and Hadith. Although many ontologies are produced to cover different aspects of Islamic topics, the production quality and quantity of Islamic resources are still in the early stages compared to other domains. This need boosts the necessity of evaluating the existing ontology toward building a resource that covers Islamic topics more comprehensibly. In this work, the data-driven evaluation method is applied to evaluate the adequacy of the available Quranic Ontologies, namely  QuranOntology and Qurany, in covering the topical concepts for one of the Islam pillars, particularly the Hajj domain. The results show that the existing ontologies do not deeply cover Islamic topics. In addition, the QuranOntology concepts coverage represents 1% precision related to Hajj terminology. On the other hand, only 14 out of 160 Hajj keywords are matched with the ontology concepts, representing approximately a recall of 8.75 %. Therefore, there is a strong need to build knowledge resources that enrich the coverage of Islamic topics.