Student academic early warning method and system based on multi-granularity task joint modeling

Abstract

The invention provides a student academic early warning method and system based on multi-granularity task joint modeling, and the method comprises the steps: obtaining student academic related data based on a school educational administration system, and carrying out the data preprocessing; performing feature extraction on the preprocessed data, representing students based on features, mapping student samples to a new feature space based on a constructive clustering algorithm, and obtaining student representation vectors; constructing a multi-task data set with courses and professions as granularities; based on the goal of minimizing empirical loss corresponding to all courses of all majors, determining a multi-granularity task joint model, and performing model training by adopting the constructed multi-task data set, wherein an objective function of the model comprises constraints on course granularity and professional granularity; and inputting a to-be-predicted student representation vector into the trained multi-granularity task joint model to obtain an academic prediction vector of the student, and further generating an early warning result of the student.

Description

technical field [0001] The disclosure belongs to the technical field of smart education and educational big data mining, and in particular relates to a student academic early warning method and system based on multi-granularity task joint modeling. Background technique [0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art. [0003] In recent years, artificial intelligence and machine learning technology have developed rapidly, and have achieved relatively successful applications in many fields such as natural language processing, image classification, unmanned driving, and medicine. With the advancement of education informatization and the concept of "smart education", machine learning technology has gradually shown great application value in the field of education. Educational data mining emerged as the times require. It is an interdisciplinary research field that integrat...

AI Technical Summary

Problems solved by technology

[0005] The inventor found that although the research in this field has been relatively extensive and in-depth, and the results obtained are relatively excellent, but due to the difficulty in obtaining data in the field of education, and the characteristics of multi-disciplinary and multi-professional education in colleges and universities, the sample size in a single data set is often relatively rare

However, the modeling method based on machine learning has certain requirements on the amount of data, and the insufficient sample size is difficult to meet the needs of training a complex and effective model, which makes researchers face huge challenges

In this case, various classification methods are not yet mature. How to construct an accurate prediction model of students' academic performance based on a small amount of student sample information still requires a large number of researchers to continue in-depth research and exploration.

Images