This project aims to develop a novel and efficient collaborative translation teaching system based on a field-sensitive corpus covering specialized translations such as mass media translation, audiovisual translation and financial translation. Its structure will be designed in such a way as to be compatible with that of Textwells, a computer-assisted, data-driven platform for general translation and bilingual learning which has been in use on several undergraduate translation courses offered by HSS. Our proposed field-specific system will be used to complement Textwells as the latter may not contain sufficient data for the teaching and learning of specific subject-related translation skills but is ideal as a platform for introductory or general training purposes. The ultimate goal of our system is, therefore, to replace the subjective impression-based or submissive machine-led teaching mode prevalent in many translation courses with an upgraded machine-assisted translation teaching model supported by focused data and the knowledge network, with a view to enhancing students’ translation knowledge and skills, and most importantly, empowering students with self-learning abilities to engage in lifelong learning in the technology era.