Improving Feedbacks for ITS Assessment of Concept Maps

Assessment in intelligent tutoring system (ITS) on concept maps (CM) matches an expert CM to a learner CM. Feedbacks are provided to the learner as semantic comments andvisual corrections. In this paper, quality of feedbacks is improved by using an ontological semantic for matching, formalized as a correlation feedback. Matchings are selected based on an overall assignment solution providing a suboptimal set of correlation feedbacks to the learner.