Navigation auf uzh.ch
The AI-assisted grading platform has demonstrated significant potential for improving assessment efficiency and consistency. While the project funding concluded in 2024, we envision several paths for potential future development and application, namely the integration with existing systems as well as a generalized research application of the approach.
Our experimental platform has validated key functionalities for AI-assisted grading. The next logical step is integrating these features directly into established e-Assessment platforms like OLAT and Inspera. This integration would eliminate the need for a separate grading platform while making AI assistance immediately available to a broader audience of lecturers.
We plan to adapt our AI-assisted grading approach for formative assessment in KlickerUZH. This integration will enable efficient grading of open-ended quiz questions, providing students with immediate feedback during learning activities. The lower stakes environment of classroom quizzes offers an ideal testing ground for further development of AI assistance in assessment.
The core technology developed for exam grading shows promise for broader research applications. By generalizing our approach, the platform could support various labeling and human annotation tasks in research contexts. The combination of AI-assisted content highlighting, similarity-based ordering, and quality assurance through clustering could benefit any scenario requiring systematic content analysis and classification.
While the project funding concluded in 2024, we remain open to analyzing specific datasets and evaluating AI functionality if financing is provided. This could include:
• Analysis of new exam types
• Evaluation of different AI models
• Performance comparisons
• Integration feasibility studies
This approach ensures that the knowledge gained from our initial implementation can continue to benefit the educational technology community while maintaining a sustainable development model.