Association Analysis-based Computational Thinking Evaluation And Its Application

Computational thinking is a way of thinking that uses computational tools and methods to solve problems.The connotation of the computational thinking,teaching method,teaching mode and so on all have the corresponding research,but the study of computational thinking evaluation at home and abroad is still in the stage of relatively weak,the existing relevant evaluation research is limited to the traditional curriculum evaluation method,therefore,the cultivation of the computational thinking effects cannot be quantified well,can't further excavate students' characteristics.How to quantitatively evaluate students' computational thinking ability is an important direction of future research in the field of computational thinking education.Unified mode of teaching has been unable to meet the different needs of each student.The premise of allowing students with different abilities to accept different training strategies is to need an evaluation system and rating model or method that can divide student groups.Therefore,this topic combined with the relevant data mining methods under the background of big data to study the evaluation model of computational thinking and student feature mining,which makes the evaluation method more objective,the evaluation results more accurate,and can better measure the characteristics of different students.To further prepare for teaching.This topic investigated the evaluation methods in the field of computational thinking,and analyzed the disadvantages of such methods,such as strong subjectivity,which is not conducive to the correlation between objective evaluation indicators,and then affect the evaluation of students' ability.The grey relational analysis model can analyze the change relationship among the objects more objectively.Therefore,based on the traditional grey relational analysis model,this paper proposes a weighted grey relational analysis evaluation model.In this evaluation model,the final weight of each evaluation index is obtained by clustering experts and combining the weight within and between expert classes.The experimental results on a data set composed of students' data collected from a certain university show that the weighted grey relational analysis model can effectively allocate the weight of each index according to the weight of experts and overcome certain subjectivity,and complete the initial classification of students.In order to ensure the accuracy of classification results,this paper proposes the mining of association rules based on computational thinking based on three-way decisions because students in each category's boundary domain are prone to misjudgment.First of all,according to the three-way decision theory,the initial classification of the students in each category three categories,respectively is divided into the domain of positive and negative,domain and edge boundaries,respectively corresponding to the corresponding strategy,maintain the initial results of positive region,to analysis the domain in the middle of the initial results,give up negative domain of the initial results,final update student classification.The results show that the three classification methods can reanalyze the students in the boundary area of each category,thus improving the accuracy of student classification.On this basis,the characteristics of students of different ability levels are excavated,and then corresponding training measures are given to students of different ability levels,so as to achieve differentiated and personalized training.