Research On Key Technology Of Education Data Verification

Education Statistics Foundation Database Platform is a tool for collecting basic identities and main attribute information of various schools.Every year,this platform collects a lot of education data,which contains rich educational information.High quality educational data is the guarantee for a government or an institution to make correct decisions.With the growing maturity of the educational system,the ability of educational institutions is different to control data quality,leading to the only a amount of accumulation and lack of quality guarantee.In the process of data acquisition and input,there are unavoidable errors and redundancies in raw data because of the influence from many factors,such as human error,environmental factors,interest factors,etc.In view of the above problems,the main work of this dissertation is as follows:1.In the process of data acquisition,the unlimited data range leads to large errors in input data.Therefore,this paper summarizes the advantages and disadvantages of the existing data verification algorithms based on the analysis of the traditional data checking algorithm,and a human-computer interaction data check algorithm is proposed.The algorithm combines the least squares support vector machine with the artificial adjustment.Moreover,it considers interaction between indicators,and then predicts Threshold interval of indicators,so as to reduce the fault data from data source.When the user collects data,this method realizes real-time data range control to achieve the improvement of data quality.2.This paper studies the knowledge of rule engine and rule base,puts forward the general model and formula of rule verification.The rule verification method is divided into two parts:rule engine checking and rule base checking.The rule engine checking mainly implements business rule checking,and the rule base checking mainly implements data verification between multiple tables.These two methods have different emphasis,which all can make the separation between program design and business logic so well,reduce coupling degree,and have a good dynamic adaptability.3.Finally,this paper uses the AHP method to determine the weight of the index,and to evaluate the quality of data through the model of five tuple M=(T,I,Ru,W,Re).A data quality control method based on "check-evaluation-feedback-optimization" is developed to control the quality of education data,to solve the problem of educational data quality effectively,and to provide high quality data for the use of educational data.This method has been implemented in the Educational Statistics Platform of a city.More than 50000 errors have been corrected,and the accuracy has increased by about 30%compared to the previous.It not only completed the construction of high quality basic database of education,but also successfully completed the data reporting task in 2017.It effectively improves data quality and provides support for government decision-making.