Intelligence Shot

The idea of the Intelligence Shot is to analyze the students' level of knowledge according to the tasks they have already completed. Gaps in knowledge can be filled with the help of the eTrainer.

An extension of the knowledge tracking model "Deep Knowledge Tracing plus" is used. In addition to the learner's task history, the knowledge concepts behind the individual modules are included in the prediction for knowledge tracing. The aim of the prediction is to model the learner's current level of knowledge by giving the model a probability for each question and level of difficulty with which the student will answer the next question correctly.

The development of the AI component is almost complete. It is already possible to predict with an accuracy of over 98% whether a student will solve the next task of a certain type correctly or not. All that is missing is a stable integration into the eTRainer and options for retraining the AI with new data.
The eTRainer had to be re-implemented due to technical difficulties. This implementation is currently in its final stages. A first test run with students has already taken place. The feedback was predominantly positive.

The development is largely complete. Now it is a matter of clarifying the final details, deploying the individual components together as an executable system and evaluating them.

From time to time, student projects are advertised as part of the eTRainer Intelligence Shot. These are often posted in the B building on the first floor or next to B208. Alternatively, these can also sometimes be found under VAPS or can be requested directly from Prof. Dr.  Haffner.

Additional information on the eTRainer can be found here.