Modeling Students' Memory for Application in Adaptive Educational Systems.

Human memory has been thoroughly studied and modeled inpsychology, but mainly in laboratory setting under simplifiedconditions. For application in practical adaptive educationalsystems we need simple and robust models which can cope withaspects like varied prior knowledge or multiple-choicequestions. We discuss and evaluate several models of this type.We show that using the extensive data sets collected by onlineeducational systems it is possible to build well calibratedmodels and get interesting insight, which can be used forimprovement of adaptive educational systems.