Evaluation of Quran Recitation via OWL
Ontology Based System
Eman Elsayed and Doaa Fathy
Mathematics
and Computer Science Department, Al-Azhar University (Girls Branch), Egypt
Abstract: The Linguistic miracle in the Holy Quran leads to many challenges in
Automate Quran recitation evaluation. This paper considers one of suggestions
of how natural language processing can benefit from using ontology. In this
paper, we proposed a general automatic system to evaluate Quran recitation
according to Hafs reading. That is via integration the ontology based as
artificial intelligent knowledge representation method and Automatic Speech
Recognition (ASR) technology as a way of interaction with computer. Our
proposed system solves the problem of evaluating all intonations (Tajweed) in
the same time in addition to evaluate set of Quran segments in the wright
arrangement of reading. The system uses Mel-Frequency Cepstral Coefficients
(MFCC) and Vector Quantization (VQ) respectively in feature extraction and
dimension reduction on Arabic speech. Also, we construct Quran ontology based
for Quranic speech and integrate it with information retrieval system. Quran
ontology based is the first version to merge Quran meaning" Tafseer"
and its recitation in the Universal oral exam to take advantage of semantic
property of ontology. Experimentally, our system gives good accuracy for Quran
recitation evaluation.
Keywords:
Holy
Quran, ontology, automatic speech recognition, feature extraction, Information
retrieval system.