Mixed-Language Text Detection , Enhancement Using NLP Techniques
This project aims to develop an intelligent system capable of processing mixed Arabic–English text using modern Natural Language Processing techniques. The system detects the language of each segment within a mixed-language sentence, translates non-target segments using neural machine translation models (such as MarianMT), and finally enhances the structure and readability of the output sentence. The project includes an interactive Streamlit interface that allows users to enter text, select the translation direction, and view results instantly.