Machine Learning and Fuzzy Logic Techniques for Personalized Tutoring of Foreign Languages

Intelligent computer-assisted language learning employs artificial intelligence techniques to create a more personalized and adaptive environment for language learning. Towards this direction, this paper presents an intelligent tutoring system for learning English and French concepts. The system incorporates a novel model for error diagnosis using machine learning. This model employs two algorithmic techniques and specifically Approximate String Matching and String Meaning Similarity in order to diagnose spelling mistakes, mistakes in the use of tenses, mistakes in the use of auxiliary verbs and mistakes originating from confusion in the simultaneous tutoring of languages. The model for error diagnosis is used by the fuzzy logic model which takes as input the results of the first or the knowledge dependencies existing among the different domain concepts of the learning material and decides dynamically about the learning content that is suitable to be delivered to the learner each time.