Using generative AI to develop assessment tools in LMS Moodle

The article examines the use of generative artificial intelligence to automatically create test-item banks from university lecture materials and convert them into Moodle XML. The study aims to assess the suitability of DeepSeek and the Perplexity hybrid platform for producing pedagogically meaningful and technically valid test materials importable into Moodle. The empirical material comprised eleven lectures from the course Infographics and Visualization in Education. DeepSeek processed seven lectures and generated 70 items, whereas Perplexity processed four lectures and generated 40 items. The analysis addressed item-type diversity, cognitive level, wording dependent on the context of a particular lecture, XML syntax, answer-weight settings, and the implementation of Cloze items. The findings show that both systems can substantially reduce the time needed to create a question bank, although their outputs require expert review. Fact-oriented items accounted for 70% of the DeepSeek sample and 90% of the Perplexity sample, producing an overall proportion of 77.3%. The DeepSeek-generated files contained errors involving fraction values, decimal separators, the structure of answer and text tags, and the content of the name tag. The Perplexity files were imported without detected syntax errors; however, they emphasized factual recall, included context-dependent references to lecture materials, and offered limited acceptance of alternative correct answers in Cloze tasks. The study concludes that technical diversity in item formats does not automatically ensure cognitive diversity. Effective use of generative AI therefore requires detailed prompts, automated XML validation, deliberate distribution of tasks across Bloom’s taxonomy levels, and subsequent pedagogical and psychometric evaluation.

1. Vikhman V. V., Mindigulova A. A., Romm M. V. Artificial intelligence in education: an overview of opportunities and limitations. Ideas and Ideals. 2024;16(4, pt. 1):167–188 (in Russian). DOI 10.17212/2075-0862-2024-16.4.1-167-188.

2. Koshkina E. A., Bordovskaya N. V., Gnedykh D. S., et al. Generative artificial intelligence in higher education: a review of theoretical approaches and application practices. Higher Education in Russia. 2025;34(6):36–57 (in Russian). DOI 10.31992/0869-3617-2025-34-6-36-57.

3. Kuzminov Ya. I., Kruchinskaya E. V., Gruzdev I. A., Naumov A. A. Lagging and leading: how students use generative artificial intelligence for educational purposes. Higher Education in Russia. 2025;34(6):9–35 (in Russian). DOI 10.31992/0869-3617-2025-34-6-9-35.

4. Zawacki-Richter O., Marín V. I., Bond M. et al. Systematic review of research on artificial intelligence applications in higher education – where are the educators? International Journal of Educational Technology in Higher Education. 2019;16(39). DOI 10.1186/s41239-019-0171-0.

5. Mustafa M. Y., Tlili A., Lampropoulos G. et al. A systematic review of literature reviews on artificial intelligence in education (AIED): a roadmap to a future research agenda. Smart Learning Environments. 2024;11(59). DOI 10.1186/s40561-024-00350-5.

6. Oliveira R., Hernández A., Garbin M., Dos Reis J., Azevedo R. Automatic question generation with large language models: a survey. ACM Computing Surveys. 2026;58(15), Article 383, 34 p. DOI 10.1145/3832773.

7. Alamoudi S., Al Khuzayem L.A., Jamal A. Optimizing automated question generation for educational assessments: a semantic analysis of LLMs with structured and unstructured ontologies. Engineering, Technology & Applied Science Research. 2025;15(3):23664–23671. DOI 10.48084/etasr.10662.

8. Kurochka K.S., Basharymau Y.S. Neural network model for auto-generating tests for students in the Moodle system based on the analysis of teaching materials. Digital Transformation. 2025;31(3):66–75 (in Russian). DOI 10.35596/1729-7648-2025-31-3-66-75.

9. Malec W. A comparative study of AI-generated and teacher-constructed language tests. Roczniki Humanistyczne. 2025;73(11):149–163. DOI 10.18290/rh257311.8s.

10. Sultanaev Ya. T., Abzalilov R. R., Vildanova V. F. Verification testology: a formalized system for evaluating test items for educational systems with artificial intelligence. Human Progress. 2025;11(9):16 (in Russian). DOI 10.46320/2073-4506-2025-9a-16.

11. Smyk A.F., Tkacheva T.M. The introduction of AI in education: conditions and trends. Baltic Humanitarian Journal. 2025;(4):177–189 (in Russian). DOI 10.46845/2071-5331-2025-4-74-177-189.

12. Al Faraby S., Adiwijaya A., Romadhony A. Review on neural question generation for education purposes. International Journal of Artificial Intelligence in Education. 2024;34:1008–1045. DOI 10.1007/ s40593-023-00374-x.

13. Du X., Shao J., Cardie C. Learning to ask: neural question generation for reading comprehension. In: Barzilay R., Kan M.-Y. (Eds.). Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics. 2017;1(Long Papers):1342–1352. DOI 10.18653/v1/P17-1123.

14. Maity S., Deroy A. The future of learning in the age of generative AI: Automated question generation and assessment with large language models. Computation and Language. 2024. DOI 10.48550/ arXiv.2410.09576

15. Kardanova E.Y. et. al. New approaches to assessment: artificial intelligence as a driver of change in education. Modern Education Analytics. 2025;5:88 (in Russian).