Creating an infrastructure for semantic search of The Holy Quran using English Wordnet

Qur’an, being the highest authoritative source of the Divine knowledge and wisdom has been widely and deeply studied, analyzed, memorized and recited by Muslims as well as non-Muslims in the past centuries.  With the advancement in the Information Technology, a number of Information Retrieval, multimedia and Natural Language Processing (NLP) tools have been developed for recital, memorization and search from the Arabic and English translated text of the Holy Qur’an. This research is an attempt to incorporate the semantics in the Qur’an search.   In this research we introduce a framework based on a variant of widely used WordNet database particularly for the English translation of the Qur’an.  The target data set comprising Chapter 2 of English Translated Qur’an has been annotated using this framework.  A front end query enhancement procedure has been developed which expands the user’s query based on the proposed framework and performs the search on the annotated text of Qur’an.  The results are promising with almost 59% precision and 59% recall for abstract user queries which do not directly appear in the text of the Qur’an in their syntax and morphology.  For non-abstract terms that directly occur in the Qur’an, the QuEST system returns results with an average of 77% precision and 88% recall which are higher than any other online tool.  A concept hierarchy of Qur’anic terms recognized automatically and semi-automatically, is a major contribution of this research.