Personalised instructional feedback in a mobile-assisted language learning application using fuzzy reasoning

This paper addresses the interesting issue of mobile-assisted language learning using novel techniques for further improving the adaptivity and personalisation to students. The domain model of the system includes English and French language concepts, and its user model holds information about students and their progress. It also embodies a database of categories of errors and misconceptions which have been reported as common in the related literature. The system is also responsible for conducting model-based error diagnosis using machine learning techniques and identifying errors such as knowledge transfer, spelling or verb mistakes. In conjunction with error diagnosis, the system employs fuzzy logic to automatically model these misconceptions and errors and then provide personalised feedback to students based on their personal learning needs. The system has been fully evaluated, using the CIAO! framework and t-test. The evaluation results are positive and encouraging regarding the educational effectiveness.