Research On The Collection And Utilization Of Student Learning Performance Data

In the past few years,the continuous advancement of information technology,data as a derivative of the application,has played a practical role in all walks of life,and in the field of education,as the educational data continues to accumulate,through the use of certain technical tools,the value of educational data is also Constantly appearing.However,some basic questions,such as what kind of data we should collect,and how we should realize the value of the data,are gradually emerging in the industry and researchers.This study focuses on student performance data and explores how to collect and use student performance data.First of all,this study summarizes the research status of student learning performance data.Most of the offline learning performance data collection adopts sociological research methods such as field classroom observation,in-depth interview method and questionnaire survey method.Online learning performance data collection relies on The recording and recording technologies such as log files and buried point records in LMS,ITS,and MOOC are mainly based on single factor or multi-factor statistical models.Based on our understanding of existing work,we are looking for current data collection and utilization deficiencies.After that,this study builds a learning performance data collection framework based on the existing theory.From the "everything is everything" thought put forward by the early natural philosophy Pythagorean school to the current big data trend,we must make changes from the conceptual change level,and quantify the "full process" of students’ learning performance,and collect "all data".The key element is to carry out the collection work.Therefore,based on the reflection of the existing theory,it is proposed to re-examine the existing work from the perspective of "human" and "space",highlighting the relationship data between the elements,mainly reflecting Relationship data between students and students,support relationship data between teachers and students,feedback relationship data between students and teachers,functional relationship data between objects and people,and transformation relationship data between people and objects And the interconnection data between objects and objects.And put forward a variety of technical tools to match,introduce new collection tools,student performance data for the core data services and other acquisition strategies.Then,from the perspective of Marxist theory of value,this study divides the value level of students’ learning performance data,and obtains the value and relationship model including three levels of characterization value,association value and decision value.Practice and technology put forward the value realization path of students’ learning performance data from the aspects of collection work,exploration work,and cost allocation.Finally,this study combines a certain amount of public data,and focuses on the multi-level value of the data.It mainly uses SEM(structural equation modeling)and supervised learning in machine learning,unsupervised learning algorithm,and dataset from multiple angles.The processing and mining are carried out,combined with certain learning theories,and the existing relationships or mathematical models are found,and the corresponding relationship between the results and the value levels is responded and explained.