Cognitive tracking method fusing knowledge association path

Abstract

The invention discloses a knowledge association path-fused cognitive tracking method, which comprises the following steps of: 1, constructing a question-knowledge point association matrix, 2, constructing a knowledge point association matrix, 3, constructing a knowledge point difficulty library, 4, calculating a skill mode, 5, aggregating, embedding and representing exercises and knowledge points, 6, embedding and representing exercises, 7, obtaining related historical exercises, and 7, obtaining a knowledge association path-fused knowledge association path-fused knowledge association path-fused knowledge association path-fused knowledge association path-fused knowledge association path-fused knowledge association path-fused knowledge association path-fused knowledge association path-fused knowledge association path-fused knowledge association path-fused knowledge association path-fused knowledge association path-fused knowledge association path-fused knowledge association path. 8, acquiring knowledge point mastering conditions of the students, 9, acquiring skill mode mastering conditions of the students, and 10, predicting future answering performance of the students. The method can start from the thinking process of the students in question making, considers the process that the students associate the knowledge points to solve the questions, integrates the knowledge association paths, and fully excavates the association relationship between the knowledge points, so that the cognitive state change of the students can be accurately and quickly tracked, and the future answering performance of the students can be predicted.

Description

technical field [0001] The invention belongs to the field of data mining, in particular to a cognitive tracking method integrating knowledge association paths. Background technique [0002] Existing cognitive tracking models can be mainly divided into three categories: (1) probabilistic models, (2) logical models, and (3) deep learning models. The probabilistic model assumes a Markov process to represent the learning process of students, and they use the unobservable nodes in the hidden Markov model (HMM) to represent the knowledge state. Logistic models assume that the probability of answering a question correctly can be expressed as a mathematical formula for the student and knowledge point parameters, they use the output of a logistic regression function to represent the state of knowledge, and use logistic regression or factorization machines to model changes in the state of knowledge. The deep learning model uses a recurrent neural network RNN ​​to simulate the cogniti...

AI Technical Summary

Problems solved by technology

If a student wants to answer a question correctly, it is not enough for the student to master the knowledge points involved in the question. The existing model does not consider the path diagram associated with the knowledge points, and cannot use the reasonable order of knowledge points to help predict the future of students. response performance, resulting in inaccurate predictions

Images