American Journal of Applied Sciences

QArabPro: A Rule Based Question Answering System for Reading Comprehension Tests in Arabic

Mohammed Akour, Sameer Abufardeh, Kenneth Magel and Qasem Al-Radaideh

DOI : 10.3844/ajassp.2011.652.661

American Journal of Applied Sciences

Volume 8, Issue 6

Pages 652-661

Abstract

Problem statement: Extensive research efforts in the area of Natural Language Processing (NLP) were focused on developing reading comprehension Question Answering systems (QA) for Latin based languages such as, English, French and German. Approach: However, little effort was directed towards the development of such systems for bidirectional languages such as Arabic, Urdu and Farsi. In general, QA systems are more sophisticated and more complex than Search Engines (SE) because they seek a specific and somewhat exact answer to the query. Results: Existing Arabic QA system including the most recent described excluded one or both types of questions (How and Why) from their work because of the difficulty of handling these questions. In this study, we present a new approach and a new questionanswering system (QArabPro) for reading comprehension texts in Arabic. The overall accuracy of our system is 84%. Conclusion/Recommendations: These results are promising compared to existing systems. Our system handles all types of questions including (How and why).

Copyright

© 2011 Mohammed Akour, Sameer Abufardeh, Kenneth Magel and Qasem Al-Radaideh. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.