Knowledge tracking method and system integrating long-term and short-term memory and Bayesian network

Abstract

The invention discloses a knowledge tracking method and system fusing long-term and short-term memory and a Bayesian network. The method comprises the following steps: establishing a cognitive data set comprising a time sequence and a long-term and short-term memory neural network; calculating a parameter group of a Bayesian knowledge tracking model of the knowledge component corresponding to thecurrent time sequence, so that the Bayesian knowledge tracking model is used for calculating the correct answer probability prediction value of the learner to the questions in the current time sequence; obtaining a loss function of the long-term and short-term memory neural network model corresponding to the current time sequence by comparing true values of whether questions of the current time sequence in the cognitive data set are answered correctly or not, so as to obtain an optimized value of the weight parameter matrix and an optimized value of the deviation parameter matrix; traversing all time sequences of the cognitive data set to obtain an optimized value of a weight parameter matrix and an optimal value of a deviation parameter matrix of the long-term and short-term memory neuralnetwork model, so that the learner can predict the cognitive state of the to-be-tested question, and learning path planning and / or knowledge graph construction of the learner are / is carried out according to prediction of the cognitive state of the learner.

Description

technical field [0001] The invention belongs to the field of knowledge tracking, and in particular relates to a knowledge tracking method and system integrating long-short-term memory and Bayesian network. Background technique [0002] Knowledge tracking is to model the learner's knowledge learning state, so that we can track the learner's mastery of knowledge points, and further predict the learner's performance in the next answer. Knowledge tracking can capture the current real needs of learners, and is the core task in learner modeling. However, due to the diversity of knowledge and the complexity of the human brain, the human learning process is complex and changeable, which is why knowledge tracking is very difficult. [0003] There are two classic solution models in the field of knowledge tracking. One of the classic models is Hidden Markov Model (Hidden Markov Model), represented by Bayesian Knowledge Tracing (BKT), which models two learning states of learners for K...

AI Technical Summary

Problems solved by technology

The advantage of this model is that the cyclic neural network is easy to train and it is easy to obtain better model prediction results; the disadvantage is that the "black box" problem of the neural network has always existed, and it cannot provide a clear semantic explanation for the model

Images