Website-based Arabic Conjugation (Sharf) Learning: ArabicMorph

Abstract

Learning Arabic morphology (sharf), particularly verb conjugation, presents a significant challenge for students due to its complexity and reliance on rote memorization. This research aims to design and develop ArabicMorph, an interactive web application to facilitate the automatic and efficient learning of sharf. The system was developed using the Waterfall model, with requirements prioritized through the MoSCoW method. It is built on the Laravel framework and integrates the Qutrub API for verb conjugation and the DeepSeek AI API for contextual translation. System evaluation was conducted via beta testing with a questionnaire administered to 17 respondents to measure user satisfaction. The results show that all core features were successfully implemented, and the system achieved a very high user satisfaction level with an average score of 91.8%, covering functionality, appearance, and ease of use. These findings indicate that the integration of rule-based linguistic technology with artificial intelligence in a modern interface is effective in overcoming the difficulties of leaning sharf. In conclusion, ArabicMorph has been successfully realized as a practical, adaptive, and user-friendly tool for independent training. The impact of this research is a digital solution that can potentially enhance student interest and understanding of Arabic morphology, providing a foundation for future development with additional educational features