自适应学习中自适应测验长度预测研究

Prediction of Adaptive Testing Length in Adaptive Learning

自适应测验是自适应评估过程中常用的一种手段,自适应测验长度预测对自适应学习的时间管理和推荐策略有非常重要的影响。对自适应测验长度预测的研究内容和现状进行分析,基于隐语义模型和认知诊断模型构建知识模块和知识点自适应测验长度预测模型,通过在大规模数据集上实验,结果表明:知识模块自适应测验长度预测模型的预测准确度较高;优化后的知识点自适应测验长度预测模型与原模型相比,预测准确度大幅提高,模型训练时间大幅减少。研究成果对自适应学习系统的实现具有重要的应用价值。

Adaptive testing is a common means in the process of adaptive evaluation. The prediction of adaptive testing length has a very important impact on time management and recommendation strategies of adaptive learning. The research content and the current situation of adaptive testing length prediction are analyzed. Then, the knowledge module adaptive testing length prediction model and the knowledge point adaptive testing length prediction model are constructed based on the latent factor model and the cognitive diagnosis model. Through experiments on large-scale data sets, the results show that the prediction accuracy of the knowledge module adaptive testing length prediction model is high;Compared with the original model, the prediction accuracy of the optimized knowledge point adaptive testing length prediction model is greatly improved and the training time of the model is greatly reduced. The research results have important application value for the realization of adaptive learning system.