Research The Student Model And Clustering On Personalized E-learning

With the rapid development of Internet technology,providing high-quality and personalized service for learners has become a hot topic.Personalized online education cored by the domain knowledge base,supported by the computer science and technology spread on the network.It is a set of education,computer science,cognitive psychology and behavioral science for the integration of application system aimed at achieving the highest teaching goal--" individualized teaching ".After studying the existing online educational resources shows that the research in providing targeted teaching strategies according to different learning basic research,learning ability,learning characteristics,interests and hobbies,and other personalized features is still weakFocusing on the online education individualized student model building and analysis,this paper completed the main work as follows:Firstly,proposed a new E-learning student model mechanism on Internet Environment,after researched and analyzed deeply the various advantages and disadvantages of the traditional student model.The original cover model and some deviation model blend became to the knowledge model;A new personality model includes three sub-models social characteristics,media characteristics and personality traits;Interpretation of the emotional and cognitive model.On this basis,we give the whole working process of student model,and give the learner characteristics database for analyzing learner features providing data.Secondly,analyzed the current clustering algorithm on data mining and improved the improved Kmeans algorithm----MKmeans algorithm based on the theory of mean shift.Then,verify the quality and the validity of the above improved algorithm with UCI’s Iris data sets and Wine data sets.The total F-measure value of the Iris and Wine are above 93%.The total F-measure value of the Iris data set than old Kmeans algorithms to enhance 8 percentage points,and the total F-measure value of the Wine data set higher 20 percentage points than old Kmeans algorithms.Thirdly,applied MKmeans algorithms to the cognitive model and knowledge model of the student model in this paper,and then analyzed the cognitive ability grouping for learners and the degree to master the knowledge for finding the learners,thereby,providing the adaptive learning materials and guides for different ability levels of learners.In this paper,guided by the new proposed personalized student model,tested through the classic test data set and compared with the existing algorithms,the improved algorithm has a better improvement in quality.The improved MKmeans algorithm can complete more accurately the student personalized classification and grouping,and division and location for the same type students providing the new ways and tools to improve the efficiency of online learning for the students.In conclusion,the proposed method of combined the personalized student model with MKmeans algorithm has a certain significance and potential application value in the personalized online education.