Developing ARET: An NLP-based Educational Tool Set for Arabic Reading Enhancement

This paper describes a novel Arabic Reading Enhancement Tool (ARET) for classroom use, which has been built using corpus-based Natural Language Processing in combination with expert linguistic annotation. The NLP techniques include a widely used morphological analyzer for Modern Standard Arabic to provide word-level grammatical details, and a relational database index of corpus texts to provide word concordances. ARET also makes use of a commercial Arabic text-to-speech (TTS) system to add a speech layer (with male and female voices) to the Al-Kitaab language textbook resources. The system generates test questions and distractors, offering teachers and students an interesting computer-aided language learning tool. We describe the background and the motivation behind the building of ARET, presenting the various components and the method used to build the tools.