Mining Learner Behavior Characteristics Based On Code Warehouse Programming Training Platform

In the practical education of computer majors in colleges and universities,educators often simply browse electronic programming assignments submitted by learners,and cannot objectively grade and correct their assignments.As a result,learners cannot get effective feedback,which greatly reduce the role of homework.In fact,there are a lot of data in electronic programming assignments submitted by learners,which hide valuable information that reflects the learning status of learners.In this era of big data education,how to make good use of these educational data and dig out the hidden value behind educational data through educational data mining technology is one of the problems faced by educational data mining workers.In the face of the current situation and problems,this article conducts data mining on the behavior characteristics of learners on the code warehouse programming training platform.The research object of this article is the code warehouse of 136 learners in college programming training.On the basis of statistical analysis,the cluster analysis method of educational data mining is adopted.Specifically,it is through cluster analysis of learners' submission and modification behaviors in the code warehouse to find the behaviors that reflect the learner's learning state in the learner's behavior attributes,which can provide improved learning for college computer professional learners and educators Good suggestions for results:discover the deficiencies in the learning status and the problems in the learning process as soon as possible,and improve the learning efficiency and performance.The main work of this paper is to design a code warehouse similarity calculation method to solve the problem of the mismatch of the number of learners submitting in the code warehouse;use the clustering algorithm to process the similarity matrix through the method of principal coordinate analysis.The experimental results of this article show that the committing and modifying behaviors of learners in the code warehouse can be used to evaluate the learning status of learners,which provides a basis for educators to give learners' performance and provide directions for understanding learners' learning.Provide basis for educators to optimize decision-making and optimize teaching methods to improve teaching quality and learners' learning efficiency and learning effect.