Abstract
This paper presents an intelligent model to read, understand and translate Arabic text content into Arabic signs. The proposed model includes four main phases, preprocessing, modeling language, translation, and sign generation phases. In the preprocessing phase, the corpora and the stop words will be employed. The language model includes morphological, lexical and syntax, and semantic analysis. This is in addition to stem, root extraction with ontological support, and number indication will be involved. Consequently, we have different features that represent the analyzed Arabic text (words' meanings, words ordering, syntactic features, number features...). Therefore, the generation phase takes place to generate the equivalent Arabic signs using the signer model. Accordingly, the deaf and hearing-impaired people are showing the generated stream of Arabic signs using video or 3D Avatar. The proposed solution uses programming corpus written in Arabic language, so the generated dictionary/lexicon has limited set of Arabic words with their meaning. After getting the content data from the course, the language model analyzes and understands the content and store it into deep structure or internal representation, consequently, the system will generate the Arabic stream signs based on the signer model.