Study On Learning Behavior Based On User Profile And Data Mining

With the greater popularity of the Internet,on the one hand,online education platforms are booming,and education informatization has entered a new era.Massive learning behavior data generated by learners on different learning platforms can be stored and recorded by the learning management system,The use of learning behavior data generated by learners can promptly and effectively provide learners with a series of applications such as early warning,learning path recommendation,course recommendation,etc.;on the other hand,computer technology is rapidly integrated with other fields,computer science technology and its applications The influence of computer science is becoming more and more far-reaching,attracting more and more people with different educational backgrounds to learn related knowledge and technology.The "Introduction to Computer" courses,as the "knowledge map" of many courses in computer majors,have a "leading" role in computer education.It lays a preliminary foundation for students' professional theory training and practical skills training,which brings an urgent need to research and design different types of "Introduction to Computer" courses.In order to study the successful experience of the design and construction of "Introduction to Computer" courses in famous universities at home and abroad,and to provide different types of learners with courses suitable for their learning characteristics,this article uses 4 courses opened on the "online education platform" in China and the United States."Introduction to Computer" courses are the research objects,based on CS 2013 to study the differences and connections between them.A comparative analysis of the teaching content of these courses from the level of knowledge domains and knowledge points reveals that there is a close causal relationship between the characteristics of the course's teaching objectives,audiences and the elements that make up them.Understanding these relationships helps design more targeted "Introduction to Computer" courses.In order to achieve the goal of personalized course recommendation,this article takes the Mit6.00 course of the ed X online education platform as an example.Through the construction of a label system for learner portraits,it integrates and screens out the characteristics of teaching interaction,course completion characteristics,learning input characteristics,and grades.The four dimensions of characteristics provide measurement indicators for the label system of learner portraits.After descriptive analysis,correlation analysis,cluster analysis,regression analysis and other data mining techniques,course learners are divided into three groups,and each The learning characteristics and behavior characteristics of class learners are explained in more detail,and the factors in the learning behaviors of learners that have the greatest impact on academic performance are identified.In order to visually distinguish different learner groups,lay the foundation for subsequent course recommendation research,construct individual and group portraits of learners,and make visual analysis,which will help teachers and students to accurately locate the four groups.Finally,according to the knowledge field and knowledge point level The analysis of teaching content makes curriculum recommendations for experiential and potential learners with high dropout rates and unsatisfactory learning effects,and provides better personalized recommendation service support.