This paper presents an interactive approach to learner diagnosis. In this approach, the effect of remedial instruction is used to inform back the diagnosis process. In intelligent tutoring system (ITS), the traditional hinting process is used after the diagnosis process. In this paper, our contribution lies in the fact that the hinting process is explicitly used to guide the student diagnosis. From a computational point of view, this approach allows the use of light weight task models. From a learning point of view, this approach allows the learner to reflect about his understanding of the domain, while he is guiding (unconsciously) a tutoring system towards a more accurate diagnosis of his difficulties.
展开▼
