Personalized test question recommendation method based on student portrait

Abstract

A personalized test question recommendation method based on student portraits aims at the objective current situation that students are easy to get lost in the current learning process and the learning state is difficult to accurately judge, a campus big data platform is constructed by integrating an existing system and developing a new system, and then the accurate portraits of the students are described. On the basis, a fuzzy cognitive diagnosis model is provided to reasonably judge the learning state attribute of the student, and the question answering condition of the student is predicted in combination with the requirements of the test questions on the learning state of the student. Based on the prediction information, a utility-based test question recommendation method is designed, and test questions with high answering utility are recommended to students. Compared with a traditional DINA method and a DINO method, the FDINA method provided by the invention not only achieves a lower root-mean-square error and a lower mean absolute error in the aspect of test question answering prediction results, but also obtains higher accuracy, a higher recall rate and a higher F1 value in the aspect of recommendation results, and in addition, the FDINA method can also effectively improve the earnings of students in answering test questions.

Description

technical field [0001] The invention relates to a method for recommending test questions, in particular to a method for recommending personalized test questions based on student portraits. Background technique [0002] With the development of big data, especially the application of artificial intelligence and big data in the field of education, personalized education has received more and more attention, but some problems have been exposed in the development process: first, the data mining of students is not enough , mainly depends on students' academic performance, and lacks data that can reflect their practical ability. These data can more accurately describe students and play a more important role in the process of personalized education; secondly, students' participation in innovation and entrepreneurship education activities It is difficult to judge the learning situation; finally, there is a lack of personalized recommendations for the students' learning process, such ...

AI Technical Summary

Problems solved by technology

[0002] With the development of big data, especially the application of artificial intelligence and big data in the field of education, personalized education has received more and more attention, but some problems have been exposed in the development process: first, the data mining of students is not enough , mainly depends on students' academic performance, and lacks data that can reflect their practical ability. These data can more accurately describe students and play a more important role in the process of personalized education; secondly, students' participation in innovation and entrepreneurship education activities It is difficult to judge the learning situation; finally, there is a lack of personalized recommendations for students' learning process, such as not being able to recommend more suitable topics to students

Both the DINA model and the DINO model are widely used cognitive diagnostic models, but both are typical discrete cognitive diagnostic models and cannot accurately diagnose students' learning status

[0004] Although the current recommendation technology has been widely studied, it is relatively lack of evaluation of recommended objects from the perspective of utility, which leads to the need for users to further select the recommended results

Images