The Research Of Uncertain Reasoning-Based Student Model Of Interests And Habits

Modern distance education is a new type of education based on traditional distance education, which is developing with the development of computer technologies, multimedia, databases and Internet. It has natural superiority and broad prospects compared with the traditional distance education.On Internet, a few sites begin to build personality pages by adopting the techniques of user model-building and personality recommendation in order to improve service quality. Rounding the personality service, researches of Internet-based modern distance education system are being done.A good distance education system should build the student's user model well and the built user model can represent the characteristics of the student. Meanwhile the system can make personality recommendation according to the user model to complete the personality education services. But building of a student user model can not do without the student's interests and habits, abilities, study rating and requirement. That requires that the system be able to analyze student's interests, error statistics, knowledge level estimating, study ability estimating and notes system etc. By that way, the system can abstract useful information from actions of students' studying.Student's user model should be flexible because courses change frequently and students' habits won't be the same at different courses. So the manual building of user model and example building of user model can't be chosen, the automatic building of user model must be the only choice. But there are some erroneous decisions exist in the automatic method. And how to overcome erroneous decisions should be a difficulty. In this dissertation, methods of uncertain reasoning including the method of natural frequency and the method of reliability are used to reduce the effect of erroneous decisions. Personality pages based on the data from the personality user model are then recommended.For the above-mentioned idea, this dissertation designed an algorithm model to arrange individuation-oriented distance education resources according to a student's actions. The key point is to use the technology of AI to count, judge and analyze the quantity of different students' clicks on all kinds of resources, to find out the pattern that exists in the actions of a student, then to build and improve the student user model and recommend personality pages based on the data from the user model.An experimental system has been built based on the algorithm model of this article. The feasibility of the algorithm model has been tested and verified from the working of the experimental system. And correlative conclusion has been worked out.