Automated Correction of Open-Ended Questions

In our P8 project "Automated analysis of open examination questions" we address the topic of (semi) automated correction of open-ended questions.

What is possible so far? Can (semi) automated correction reduce the time and effort to correct open-ended questions? Then grading an exam with hundreds or even thousands of text-answers can be time consuming. The AI-supported platform supports human graders in their work.

It improves consistency and objectivity which allows for a fairer and more transparent grading process which eliminates potential biases that may arise from subjective interpretation or personal preferences. Grouping similar answers, highlighting keywords, predicting the quality of answers, or detecting cheaters are further elements the platform offers.