Learner's knowledge modeling using annotation and Bayesian network

Abstract

Learner's knowledge assessment is very important in the e-learning system. Knowledge assessment is effective for knowledge gap discovery. Knowledge gap, Causes the learners do not understand educational content correctly. This paper presents a new method for learner's knowledge modeling based on knowledge gap discovery for concepts of educational content. There are two methods for learner's knowledge gap discovery: 1- Explicit Method 2- Implicit method. The explicit method is based on a questionnaire. In this method directly asks about various concepts of educational content from learners. Learner's answers show the level of learner's knowledge and knowledge gap regarding each concept. But, in the implicit method, knowledge gap discovery is done without direct questioning. In this paper, implicit method has been used by annotation. Annotation provide a way for learners to present their ideas and issues directly through comments, questions, and other reactions when learners as read. The main aim of this work is modeling knowledge and the knowledge gap of any learner to the concepts by Bayesian networks. The test project is done for 25 students in three fields (E-commerce, Computer, other) in three degrees (bachelors, master, PhD). The proposed method is evaluated so that the pre-test will be held for learners and the result of the pre-test is compared with the predicted knowledge.