Quran ‘Search for a Concept’ Tool and Website
The title is a dissertation by Noorhan Hassan Abbas introducing a web tool for building ontological constructions for Quranic themes. It was submitted for the degree of Master of Science by Research in the University of Leeds, School of Computing, June, 2009. The following is an excerpt of its abstract.[1]
Objective of the dissertation
The Quran is undoubtedly an important book, covering many themes and concepts. To help readers of the Quran to find and understand these concepts, this project developed a bilingual (English/Arabic) comprehensive search tool for the Holy Quran that significantly enhances recall and precision when searching for concrete concepts as well as for abstract concepts.
Introducing the tool
As most of the available tools on the web use one English translation in the search process, the average recall and precision values for these tools are 54% for recall and 48% for precision whereas this tool proves 87% recall value and 58% for precision. Besides, it is the only tool that offers users a comprehensive hierarchical classification of Quran abstract topics or themes using expert knowledge imported from ‘Mushaf Al Tajweed’ ontology of the Quran. The index of topics covers nearly 1100 concepts in the Quran.
Concepts in the project
The concepts in this index have an aggregation relationship; the hierarchy of concepts is non-reflexive, non- symmetric, and transitive.
Discription of the tool and process
Version 1 (V1) of the tool discussed in this thesis consists of two modules: the keyword search module that searches for concrete concepts and a comprehensive index of topics module that covers all the abstract concepts. The keyword search tool extends keywords to synonyms and uses eight parallel English translations of the Quran that have been stemmed and tokenized as well as the original Arabic. This proves to enhance the results of the search significantly. Also, extending the search process to look for lemmas and morphemes rather than just the keyword increases the accuracy of the search. The abstract concepts module is a list of all the themes covered in the Quran corpus which consists of fifteen main concepts that subdivide into sub-concepts and sub-sub concepts covering all the themes of the Quran.
This tool was published on the WWW (version 2 or V2) using Google Appengine SDK and server, in addition to the Yahoo! User Interface Library. Another web tool (version 3 or V3) for the Quran concepts was developed which enables users to search for concepts (concrete and/or abstract) using the powerful google search engine.