Adaptive Learning Support

In many domains, students need to learn how to solve a problem in multiple steps. For example, Gaussian elimination, performing a hypothesis test, and constructing a Java or Haskell program all require a number of steps to complete a task. At each step, a student might make an error, or might not know how to proceed. In an online environment in which a student can practice with such tasks, it is important to offer support at the level of steps. We have developed Ideas, a framework that can automatically generate feedback and hints in stepwise exercises, for any domain. We have used the framework to support solving tasks in the domains of high-school algebra (solving equations, and much more), logic, statistics (hypothesis testing), programming, and more.

More info: http://ideas.cs.uu.nl; https://ideastest.science.uu.nl.

Textbooks are high-quality information resources written by experts, organized and formatted to facilitate their understanding. As a result, their formatting and structural elements carry various semantic signals encoded by their authors. When extracted and formally represented as a knowledge model they can support reasoning and inference, linking and restructuring, enrichment and personalisation.

More info: https://intextbooks.science.uu.nl.

Open student modelling has been known to have several positive effects on learning, with the help of adaptive systems including increased motivation, support of self-reflection and self-regulation, fostering trust of students in the system’s ability to help them etc. Yet, more research is required to methodically analyse the relations between different properties of open student models and potential improvement that they can cause in various contexts. What the model should communicate back to the student (activity, progress, success?) how should it present it, how the interaction between the model and the student should be structured, what is the purpose of opening the model (merely presenting the current state, or attracting attention of the student to problematic components, or navigating the student to the remedial actions/content) are among the questions to be explored. Another interesting challenge is to effectively transfer the interaction between a student and an open model to the mobile platform that becomes the go-to context for working with educational content.