Knowledge measurement-oriented test question, knowledge and ability tensor construction and labeling method

Abstract

The invention belongs to the technical field of education data mining, and discloses a knowledge measurement-oriented test question, knowledge and ability tensor construction and labeling method. In combination of a Q matrix and Bloom cognitive field education target classification, knowledge point mastering is divided into six cognitive ability levels: knowing, understanding, application, analysis, integration, evaluation and test question, knowledge and ability tensor construction. An interpretable test question label prediction model is constructed by adopting an active learning strategy, interpretable label prediction information entropies are obtained, unlabeled samples are input into the prediction model by utilizing the trained interpretable test question label prediction model, andthe label prediction information entropies with relatively high interpretability are fed back, so that man-machine coordination is carried out. According to the method, the influence of subjectivityof manual labeling on the TKA tensor is reduced, the labeling accuracy and efficiency are high, and the expert labor cost is greatly reduced. The method is high in mobility, can be applied to examination and annotation of test question knowledge points of all subjects, and is better in applicability.

Description

technical field [0001] The invention belongs to the technical field of educational data mining, and in particular relates to a test question, knowledge and ability tensor construction and labeling method for knowledge measurement. Background technique [0002] Currently, the closest prior art: In the field of education, the diagnostic assessment of an individual's knowledge structure, cognitive processing skills, or cognitive process is usually called a cognitive diagnostic assessment. Knowledge or cognitive processing skills are collectively referred to as attributes, and learners' knowledge structure and cognitive processing skills are latent variables that cannot be directly observed. In the evaluation theory of cognitive diagnosis, attribute is the most basic concept, which refers to those potential and implicit psychological characteristics that affect the external behavior of the subjects. In a teaching test, the subject is the learner, whose external behavior is refl...

AI Technical Summary

Problems solved by technology

However, in most cases, the project parameters obtained from the wrongly marked Q matrix at the beginning have large errors, which will seriously affect the accuracy of the Q matrix estimation results based on the project parameters

[0013] The third is to estimate the Q matrix based on the learner's response data. This method is mainly divided into non-parametric estimation method and parameter estimation method. lower sex

[0015] (1) In the prior art, for the characterization of the relationship between subject knowledge and test questions, the current research mainly focuses on the representation of the relationship between learners-test questions and test questions-knowledge, and does not comprehensively consider and represent the relationship between test questions-knowledge-ability , resulting in the failure of the follow-up diagnostic model to obtain effective support from the data layer, and unable to diagnose the learner's knowledge mastery and knowledge cognition level

[0016] (2) The traditional learning diagnosis is based on the learner’s answer response matrix and the test question-knowledge point correlation Q matrix to diagnose and analyze the learners. It can only judge whether the knowledge points are mastered, but the degree and level of knowledge mastery cannot be estimated. And provide less domain information for interpretability modeling

[0017] (3) In the existing technology, most of the labeling of test questions is done manually, but in the online learning platform, the number of unmarked test questions is huge. If the test questions are marked by education experts in the subject field, the labor cost will be high. Long labeling time and low accuracy directly lead to poor diagnosis of subsequent test questions and inaccurate parameter estimation

Images