Research On The Application Of Data Mining Technology In University Wisdom Aid Financially

During the 13th Five-Year Plan period,with the goal of not letting a student drop out from school for the reason of family financial difficulties,China's subsidy policy has basically achieved full coverage of the education stage.The requirement of student aid work in higher education schools is to realize the full coverage of the aid policy for students with financial difficulties in families and to achieve the effect of precise sponsors.However,in the identification of poor students in colleges and universities,most of them still adopt the traditional methods through family situation questionnaire or one-card data.The acquisition of assessment information may be fraudulent,and the degree of family poverty cannot be quantified."False poverty"and”invisible poverty"cannot be controlled with the changes of family economic situation,and the effect of subsidized education cannot be tracked.Therefore,the focuses of current research on university funding include achieving the goal of precise poverty alleviation,objectively reflecting the effect of financial aid and education,and eliminating"false poverty"and"invisible poverty" as far as possible.The arrival of the era of big data has provided a new and more scientific way for college students'financial aid work,and has brought a new opportunity for higher education schools to use big data to promote rapid,accurate and efficient financial aid work.The purpose of this paper is to build a school's intelligence subsidy system by relying on school's big data platform for the research and discussion of intelligence subsidy.This system is not an identification system for poor students in a simple sense.It is based on the big data platform of the school and data mining technology.The system covers the functions of online application for subsidized projects,prediction of poor students,identification of false and invisible poverty,and analysis of the effect of subsidized education.Through collecting and summarizing the data of students'dormitory door prohibition system,students' consumption data,students' borrowing history data,students' scores in school,students' source loan data,students' family situation score and so on,after data cleaning,some unnecessary data are removed and stored in the database.Considering all kinds of factors related to poverty,five types of feature clusters are selected:consumption feature cluster,behavior trajectory feature cluster,grant and loan feature cluster,family background feature cluster and social relationship feature cluster.These help to find out the characteristics and association rules of poor students.Then we use some common data mining algorithms such as logistic regression,network nerve,random forest to get the mining model of poor students,false poor students,hidden poverty mining model,and the mining model of financial aid effect.Finally,we use the evaluation results of poor students in previous years to evaluate the accuracy of the data mining model and prediction results.The intelligent subsidy system discussed in this paper is different from the general identification system for poor students.It is not a simple one-time identification system for poor students using the campus consumption card data.The intelligent subsidy system uses all available data sources in the campus to build a mining model and predict the poverty ranking of students in a class or a major,and the ranking is not the same.It will not terminate because of the end of the identification procedure for poor students,but continuously and dynamically detect the data of poor students.When there is abnormal consumption and behavior in a short period of time after receiving assistance,reminder emails will be issued to managers.And every three months,it gives an analysis of consumption,academic achievement and award-winning of impoverished students,which objectively reflects the effect of university funding in educating people.This system can avoid all kinds of disputes in the identification of poverty students to a great extent.It is conducive to the smooth and orderly management of college students.It can also evaluate objectively whether the method of university funding for education is timely and effective.