Student cognitive diagnosis method

Abstract

The invention discloses a student cognitive diagnosis method, which comprises the following steps of obtaining the historical answer information of a student, and extracting a test question text and the predefined knowledge points contained in the test question text; calculating a knowledge point relevancy vector of each test question according to the test question text and the contained predefined knowledge points; taking the set student parameters and the test question parameters containing the knowledge point relevancy vectors of the test questions as the input of a cognitive diagnosis model constructed by a neural network, fitting the answer results, obtaining the knowledge point mastery degree vectors of the students through training, and completing the cognitive diagnosis of the students.

Description

technical field [0001] The invention relates to the technical field of educational data mining, in particular to a method for diagnosing students' cognition. Background technique [0002] Cognitive diagnosis is a very important part of intelligence education, and it is an improvement and perfection of traditional examination and evaluation. Coarse-grained student evaluation methods are usually used in traditional education scenarios, such as assessing students' abilities through students' test scores or correct answers to questions, but this alone cannot know the reasons why students answered wrong questions, nor can they be accurate. Find out where the students' knowledge gaps are. In response to this problem, student cognitive diagnosis aims to obtain the degree of students' mastery of specific knowledge points hidden in it through the analysis of students' answer records. [0003] Cognitive diagnosis is one of the key technologies in the current personalized intelligent...

AI Technical Summary

Problems solved by technology

[0009] Although the above cognitive diagnostic methods have demonstrated their effectiveness in past applications, they are linear combinations between student vectors and test item vectors, and are often unable to accurately model the complex relationship between students and test items.

Although the student vectors obtained by some methods (IRT, matrix decomposition, etc.) reflect the abstract characteristics of students and can be used to predict the results of answering questions, they have no practical and interpretable meaning, and it is still impossible to know the students' mastery of each specific knowledge point. Therefore, it is not conducive to the understanding and further utilization of diagnostic results

In addition, the design of the function mapping of the specific vector combination method (such as logistic regression in IRT, dot product in matrix decomposition) requires more professional knowledge, which is very time-consuming and labor-intensive, and the types of test questions that are usually suitable for designing functions are very difficult. limited

Images