Design And Research Of Educational Big Data Statistical Analysis Platform

Traditional statistical analysis of educational data needs to rely on manual work.The use of computer technology in the statistical analysis of educational data can reduces the waste of human resources and improves the efficiency and accuracy of statistical analysis of educational data.At the same time,education data grow exponentially and the data type is becoming more and more diversified.The emergence of big data have brought solutions to the storage and processing of massive education data.Based on the actual needs,this paper designs and develops an education big data statistical analysis platform,and conducts statistical analysis and prediction research on data of students and teachers from multiple dimensions,providing quantitative basis for management decisions,and conducts relevant research on the migration,storage and query needs of education big data.The specific work is as follows:In order to realize the automation of data processing in education bureau of a city and improve the operation efficiency and scientific decision-making,an education data statistical analysis platform was designed.Firstly,combining with the actual situation of the platform,the storage mode of structured and unstructured education data was designed reasonably.Secondly,the functions of yearbook management,educational GIS map,student migration,data analysis and data prediction were designed according to the needs of the education bureau.Finally,the data prediction model based on RNN and LSTM was studied,which can be used to analyze and predict the data of students and teachers,etc.,with certain effect,meeting the needs of a city education bureau and putting it into use.In order to meet the demand of the rapid growth of education data,an education big data platform based on Hadoop was studied and designed.Firstly,an educational data warehouse based on Hadoop was designed from the conceptual model,logical model and physical model.Secondly,the storage of educational multidimensional data model was designed and studied.Then,Based on educational big data,the comparison between HBase and Hive query efficiency and storage space in Hadoop ecosystem is tested.HBase+Hive integrated architecture was adopted to complete the statistical analysis task of educational data,and the function of educational data statistical analysis platform was transplanted to the educational big data platform based on Hadoop.Finally,a performance test was conducted on the conversion efficiency of educational big data in the ETL link,demonstrating the effectiveness of the educational big data platform based on Hadoop.