Ontological Approach for Semantic Modelling of Malay Translated Qur’an

This thesis contributes to the areas of ontology development and analysis, natural language processing (NLP), Information Retrieval (IR), and Language Resource and Corpus Development. Research in Natural Language Processing and semantic search for English has shown successful results for more than a decade. However, it is difficult to adapt those techniques to the Malay language, because its complex morphology and orthographic forms are very different from English. Moreover, limited resources and tools for computational linguistic analysis are available for Malay. In this thesis, we address those issues and challenges by proposing MyQOS, the Malay Qur’an Ontology System, a prototype ontology-based IR with semantics for representing and accessing a Malay translation of the Qur’an. This supports the development of a semantic search engine and a question answering system and provides a framework for storing and accessing a Malay language corpus and providing computational linguistics resources. The primary use of MyQOS in the current research is for creating and improving the quality and accuracy of the query mechanism to retrieve information embedded in the Malay text of the Qur’an translation. To demonstrate the feasibility of this approach, we describe a new architecture of morphological analysis for MyQOS and query algorithms based on MyQOS. Data analysis consisted of two measures; precision and recall, where data was obtained from MyQOS Corpus conducted in three search engines. The precision and recall for semantic search are 0.8409 (84%) and 0.8043(80%), double the results of the question-answer search which are 0.4971(50%) for precision and 0.6027 (60%) for recall. The semantic search gives high precision and high recall comparing the other two methods. This indicates that semantic search returns more relevant results than irrelevant ones. To conclude, this research is among research in the retrieval of the Qur’an texts in the Malay language that managed to outline state-of-the-art information retrieval system models. Thus, the use of MyQOS will help Malay readers to understand the Qur’an in better ways. Furthermore, the creation of a Malay language corpus and computational linguistics resources will benefit other researchers, especially in religious texts, morphological analysis, and semantic modelling.