Research On Design And Key Technology Of Academic Early Warning System Based On Hadoop

With the expansion of colleges and universities,the number of students has increased dramatically.The academic and behavioral data of a large number of students deposited in schools for many years have not been effectively utilized.By studying student behavior and school behavior data,using big data technology and data mining technology to construct the relationship model between students' academic and behavioral data,and based on this,design the student's academic early warning system.Through the research and development of the system,on the one hand,it can help students change their learning behavior,improve their academic performance,and complete their studies on schedule.On the other hand,it helps teachers to fully understand the academic status of students,optimize the curriculum system in a targeted manner,and promote the process of school education reform.Therefore,how to integrate the multi-heterogeneous data in combination with student behavior and academic characteristics,design and develop a personalized student academic early warning system is a problem worth studying.Based on the existing theories and methods,this research starts with the multi-source heterogeneous big data processing technology based on Hadoop,and starts from the multi-source heterogeneous data extraction,dimension reduction,cleaning and other pre-processing techniques,and designs the academic early warning system.The overall business process,the overall technical architecture and the construction of a big data processing platform.Secondly,based on the characteristics of data,draw on the idea of data mining,driven by big data,analyze the student's performance data,Internet data and consumption data,and use clustering algorithm to identify and analyze the abnormal points.On this basis,the relationship model and information and feedback mechanism between students' online and consumer behaviors and learning status are constructed,and the student's academic early warning module is designed.Finally,the three elements of software engineering methods,tools and processes are combined,based on the structural paradigm.From the aspects of software architecture design,database design,system function module design,etc.,the prototype system of student academic early warning is realized.