A recommendation method for financial aid for poor students based on multidimensional analysis of school behavior data

Abstract

The invention relates to a recommendation method for impoverished students based on multidimensional analysis of school behavior data, which solves the defect that it is difficult to accurately recommend impoverished students to be aided compared with the prior art. The invention comprises the following steps: acquisition of historical behavior data; feature extraction of historical behavior data; training of recommendation model; acquisition of behavior data to be analyzed; feature extraction of behavior data to be analyzed; Based on the in-school data generated by students, the invention extracts features of multiple dimensions, uses these features to establish a classification model, and accurately judges the poverty situation of students and makes decisions by means of the classification model.

Description

technical field [0001] The invention relates to the technical field of big data analysis, in particular to a method for recommending financial assistance for poor students based on multidimensional analysis of school behavior data. Background technique [0002] The advent of the era of big data has provided new ideas and technical support for the funding of poor students, and has also brought new opportunities for colleges and universities to use big data to promote fast, convenient, efficient and accurate funding. Using big data mining and analysis technology and mathematical modeling theory to help managers grasp the real behavior patterns of students during school, discover "hidden poverty" and suspected "false identification" students, and achieve precise funding. [0003] At present, the precise subsidy for poor students in school is still in the exploratory stage. There is no unified evaluation method in China. There is no unified demarcation standard for poor students...

AI Technical Summary

Problems solved by technology

[0003] At present, the precise subsidy for poor students in school is still in the exploratory stage. There is no unified evaluation method in China. There is no unified demarcation standard for poor students in school. The subsidy for poor students lacks systematic and standardized management. It is very cumbersome and causes a lot of waste of data resources

Although some technologies have put forward some viewpoints and ideas, none of them can meet the practical application or are difficult to realize. For example: the patent application document with the patent number 201710223971.6 and the patent name is the method for predicting students' poverty status based on data mining

Although it is aimed at analyzing the data of students in school, it directly uses the big data platform hadoop and spark, and the model uses random forest, and does not carry out targeted technical data classification for the data of students in school, so that the classification results are not accurate. ideal

Images