Arabic is the first or second language to more than a billion human beings. A majority of Arabs are Muslims who consider the Holy Quran as the word of God and the main source of guidance and rules. In Arabic language we are still in the phase of syntax-based search, where the retrieved results are still based on the exact words the user enters. Very few approaches have been developed to move beyond that, but most of the works are not general enough to serve the majority of Arabic speakers who are attempting to find information in the Quran. Question answering is the process of retrieving answers to questions users post in a natural language. In this paper, we use semantic approaches to represent Quranic content and interpret user questions written in Arabic, and to try to find the best answer to questions from the semantic representation of the Quran. We achieved a usable model that can answer questions from the Quran even if the exact words the user entered are not present in the Quran. We implemented a prototype of our approach and the results we present later in the paper are very encouraging.
Publication Date
12/10/2014
Publisher
2nd International Conference on Islamic Applications in Computer Science and Technologies (IMAN 2014)
سمير الترتير | Samir Tartir | 4445