Student Modeling And Application Research Based On Cognitive Diagnosis

In recent years,information technology has continuously penetrated into every as-pect of human's daily life,it bringing us great convenience as well as changing our way of life,In addition to the changes brought about by traditional Internet service such as shopping,travel,entertainment and communications,it is also gradually affect-ing the current education model.From the Internet to the mobile Internet,in addition to traditional off-line education,students nowadays have come up with a brand new teaching model which is called computer-based learning.This kind of new teaching and learning model,which breaking the through time and space constraints,reduces the cost of education and provides open-ended courses and learning experiences to a lot of students.However,with the continuous development of on-line education platforms and the continuous advancement of educational informationization,a large amount of complex educational data has been generated and accumulated.How to use these edu-cational data to fully understand and discover hidden information,and how to use this information more wisely,has become one of the important research topics in the field of educational data mining.Educational data mining(EDM)is an emerging research field,and many scholars and researchers have proposed a lot of student modeling methods and applications.As a commonly used student modeling method,cognitive diagnosis provides a set of diag-nostic models for diagnosing students' learning status.However,there are still many areas to be studies and improved in the current student modeling and application.First,for the research field of student modeling,the popular cognitive diagnosis models are mostly aimed at modeling student learning status base on the record collected from a single test,and thus lacking the correlation of multiple independent student model-ing work and commonality between similar tasks,to generate accurate student learning status.Second,for the application of modeling results,there is a lack of an effective combination of cognitive diagnostic results and traditional methods.Therefore,aiming at the above research problems,this paper studies the student modeling and application based on cognitive diagnosis method.The major work and contributions are as follow:1.Proposed a multi-task cognitive diagnostic framework MT-MCD.This frame-work can diagnose more accurate and comparable student learning status by associating multiple independent student modeling tasks together.2.Proposed a personalized question recommendation method PMF-CD.This method incorporates students,learning status in the traditional recommendation model so that each students' cognitive level could be effectively and interpretably performed to gen-erate personalized question recommend.3.For the above two research,a large number of experiments were performed on real-world datasets.The experimental results prove the validity of the cognitive diagnostic framework and the personalized question recommendation method proposed in this paper.