A Study Of Educational Assistive Technologies Based On Cognitive Diagnosis Theory

With the expansion of the information technology to educational field, a variety of intelligent tutoring systems(ITS) are developing rapidly and have attracted a large number of users. ITS now has become a kind of mainstream learning environment for users’knowledge construction and collaborative learning. Compared with the traditional education, ITS is not bound by any restrictions like time or space, as a result, its autonomy and openness makes students’ learning and teachers’ teaching more efficient. But essentially, ITS is still a kind of knowledge cramming education with lots of resources. ITS is now faced with challenges such as lack of individualized teaching, shortages in flexibility and insufficient attention to students at risk. Hence, how to promote students’ academic achievement by enhancing students’ self-understanding, raising students’ learning efficiency and positivity desiderates to be resolved. Based on the increasing education-related data, it has already become the development trend of educational data mining techniques to provide effective support for education.Various educational assistive technologies have been studies in the domain of data mining and education psychology. However, there is still room for improvement. Firstly, in terms of student modeling, existing cognitive diagnostic models can’t model students with continuous skill proficiency, which cannot provide students with more detailed feedback and more valuable suggestions for improvements. Secondly, educational assistive technologies based on cognitive diagnosis till face great challenges:on the other hand, the features used in collaborative learning team formation are coarse-grained, and the efficiency of team formation algorithms needs to be enhanced; on the other side, for students’ achievement trend prediction, the features or models used are lack of interpretation which greatly limits its application on educational assist. To solve the above problems, this thesis developed a series of researches about educational assistive technologies. The major work and contributions are as follows:1. Proposed a cognitive diagnosis model SoftDINA. As an improvement of the existing diagnosis model DINA, SoftDINA automatically quantifies the students’skill proficiency in continuous and probabilistic values.2. With students’ skill proficiency getting by SoftDINA, this thesis introduced cognitive diagnosis to the problem of collaborative learning team formation, and proposed a framework for students’ collaborative learning team formation based on cognitive diagnosis. This thesis proposed two objectives of Dissimilarity and Gain, and then effective algorithms were designed to generate collaborative learning teams.3. Proposed a framework for student achievement trend prediction based on cognitive diagnostic theory. This framework applied two kinds of educational psychology theories, IRT and DINA to analyze students’ ability and skill proficiency, then several classification algorithms are applied based on student modeling results to predict students’achievement trend.