Research On Personalized Elective Course Selection Based On Improved Collaborative Filtering Algorithm

With the rapid development of information technology, the education managementsystem in high schools has changed correspondingly."The decision of the CPCCentral Committee on the reform of the education system" indicated clearly that thereis need to increase electives and reduce the required subjects, and to implement creditsystem and double degree system teaching, which the elective system is the core ofthe credit. Every university provided students with numerous and various electivecourses which the construction of course category and major, such as the organizationand management of the Curriculum Resource Center and election mode of courses isdeficient. So now it is difficult for students to choose the suitable course benefitingpersonal profession development and fulfilling individual needs. And manyuniversities implemented an incomplete credit system with the result that the studentsdo not have much choice in the course election. After all we will apply thepersonalized recommendation technologies to the elective system that can providestudents with reasonable, scientific and personalized elective recommended accordingto the needs, interests and learning preferences of students, which make the studentsavoid blindness election and improves the utilization of curriculum resources.In this paper, we mainly in-depth study the collaborative filtering technology anddesign the elective recommendation system, and the specific research works asfollows:Firstly, analyzing the advantages and disadvantages of the personalizedrecommendation technologies and putting forward collaborative filtering algorithmbased on the characteristic properties of the course and attributes-value preferencematrix. In order to achieve the real-time course recommended, we adopt thecharacteristic properties of the course and attributes-value preference matrix for theproblem of sparse data and cold start, and calculate the similarity by offline. Secondly,we build a system framework with personalized recommendation, rankingrecommendation and new courses recommendation. It is practical to reduce the error of courses recommendation, and improveinstantaneity of recommendation by constructing a elective system with personalizedrecommendation which make use of characteristic properties of the course andattributes-value preference matrix, thereby, expanding the horizons of students,increasing the autonomy of students learning and culturing innovative thinking ofstudents.