Research And Application Of Academic Evaluation Method For Computer Science Courses In High School Based On Cognitive Diagnosis

In computer science courses,the traditional academic evaluation focuses on the student's learning outcomes,which seldom reflects the development of students' cognitive level.There is only some feedback information like “scores” or “right/wrong” to reflect the students' learning effectiveness at a certain stage.It is too difficult to pinpoint the exact area where the student has difficult learning new knowledge points or skills;which infers that the traditional evaluation lacks effective diagnostic information.Cognitive diagnosis is a combination of technology and measurement,which can accurately diagnose the student's knowledge gaps,cognition levels,and knowledge structure;providing a new idea for the next-generation evaluation of computer science courses.Based on this,this study proposes a non-parametric multi-level scoring cognitive diagnosis method based on the characteristics of computer science courses and data.The EW-KNN(E-weight K-Nearest Neighbors)diagnosis method introduces distance information and has a dynamic K value.The Monte Carlo simulation method was used to test the performance of this new method.The results indicated that the EW-KNN diagnosis method has a higher accuracy rate in a small sample data,which is more suitable for computer science courses.Furthermore,based on the EW-KNN diagnostic method,this study selected the chapter of "Basic Conversion" from high school as the diagnostic content to carry out the academic evaluation of computer science.The experimental content includes three parts: test question preparation,diagnostic analysis and report design.Diagnostic analysis results show that the diagnostic information of EW-KNN has the characteristics of fine-grained and multi-dimensional compared with traditional tests.In terms of information granularity,it can be detailed to knowledge points and test questions,which overcomes the shortcomings of traditional examinations that only reflects test question scores.In the information dimension,it can correctly analyze students' cognitive structure,cognitive level,and mastering degree of knowledge points.EW-KNN solves the difficulty in distinguishing the differences among students attained the same score using traditional tests.Besides,based on the characteristics of the EWKNN method,a diagnostic report is designed from both an individual and a class perspective to provide diverse information to teachers and students.It is verified that the diagnostic report has certain practical teaching significance from the interview with three teachers.