Application Of Aprie Improved Algorithm Based On Data Mining Technology In The Identification Of Poverty Students In University

Now our country has achieved rapid economic development, but still has some poverty students due to financial difficulties can't go to university. With the increasing of the number of poverty students in universities and projects, as well as the improvement of social management work in university, education management is required to move forward with the Times. Universities are starting to accelerate the construction of campus digitalization, information construction. At present, all universities have set up their own management systems, covering students' enrollment status, personal files, grades, the party, financial management, document management, libraries, canteens, etc. These different management systems have a large amount of data information. If we can make full use of these data and information, the poverty students identified will be more effective to carry out.This paper uses data mining technology, in view of the IC of the existing water consumption data mining analysis, to find out the potential consumption characteristics of students, and then analyze their economic conditions, consumption is defined as a group of poverty students. It has an important influence on the assistant identification of poverty students. The paper uses association rules and decision tree for mining algorithm, puts forward a new decision tree based on association rule mining algorithm, applied to the cognizance of the poverty students. First of all, the analysis of the current situation of poverty students work at home and abroad; Second, learn the related knowledge, data mining association rules algorithm and decision tree algorithm are analyzed; Again, the association rules algorithm is combined with decision tree structure, generating a series of rules, extracting the characteristic structure of new properties, the restructuring data set; Finally, the algorithm to classify a set of tests, the processes by which shows the improved algorithm.The paper has chosen three poverty students' data records from different majors, a total of 856 records, to inspect the classification effects of the model, the paper makes a comparison between the final classification results and the results of the school identification, the result is consistent with a record of 87.03%. According to the mining results show that the proposed algorithm for the determination of poverty students has important significance in the future.