The use of Arabic linguistic resources to develop learning applications

Abstract

This paper aims at explaining how Arabic linguistic resources are generated and exploited to enhance Arabic acquisition. We have adopted the root and pattern approach to generate our resources using the linguistic NooJ platform. This work has been carried out in two phases: generating the linguistic resources and developing the application that exploits the pre-built resources. First, we have generated three different resources: comprehensive verbs and masdar resources linked to each other. A nouns-and-adjectives resource, where nouns and adjectives are linked to their broken plural forms. NooJ calls these resources to apply linguistic analysis to a given corpus and returns detailed annotations, which provides accurate morphological, syntactical, and semantic properties of each analyzed word. We have also used the mixed nature of Arabic masdar to implement transformational rules, which generate nominal sentences from verbal ones and vice versa. Second, we have developed an application that provides valuable learning functionalities, like full/semi verb conjugation, the extraction of broken plural forms of a given singular form, the extraction of masdar forms of a given verb form, and return words that share the same root. The developed application can be used by teachers, students/learners, and computational linguists interested in Arabic acquisition.