Computer-assisted i‘raab of Arabic sentences for teaching grammar to students

Abstract

The end-case analysis of Arabic sentences is one of the keys to their meaning. This process is called i‘raab, a daunting task for the students. The outcome of the analysis is twofold: (a) placing a proper diacritical marking on the end-cases of individual words, and (b) providing a logical justification. Our objective is to generate a full i‘raabof the sentences, which could be incorporated in a computer-assisted Arabic grammar learning environment, teaching and helping the students with the i‘raabprocess. Our system comprises four components that work together to generate a proper i‘raabfor the sentence. We devise an enhanced context-free grammar (eCFG) that covers all the cases and rules taught in Saudi schools’ grammar textbooks (grades 7–12). Moreover, our eCFGgrammar eliminates the need for specialized grammars, e.g., head-driven phrase structure grammar and link grammar, to resolve complex cases involving dependencies. Furthermore, we utilize ontology to determine the correct semantics. The system was tested on 300 sentences varying in complexity from the Arabic grammar textbooks used in the schools. Our system achieved an overall accuracy of 88.33%.