Elamur Research⌕
☰
home / papers / 2026 / Evaluating Adaptive Personalization of Educational Readings with Simulated Learners

Evaluating Adaptive Personalization of Educational Readings with Simulated Learners

Paper · 2026 · Ryan T. Woo, Anmol Rao, Aryan Keluskar et al.

Summary

Tailoring course readings to what each student already knows or misunderstands is hard for instructors to do at scale. Testing adaptive reading policies directly in classrooms is also costly. The authors build an offline framework that screens such policies with simulated learners grounded in cognitive theory. From Wikibooks open textbooks, an LLM-assisted workflow proposes learning objectives (LOs) and knowledge components (KCs). Humans then refine these in a browser-based editor called Ontology Atlas. Textbook chunks are tagged with these entities by an LLM-as-judge pipeline, and aligned readings and multiple-choice items are generated. Simulated readers absorb passages through a Construction–Integration-style memory model. They differ along DIME reader factors, revise misconceptions following KReC, and are affected by New Dale–Chall readability. Answers are chosen by scoring options against each learner's explicit memory, and BKT estimates drive which readings come next. Adaptive cohorts clearly beat matched controls in computer science, gained modestly but inconclusively in inorganic chemistry, and showed no benefit in biology. The framework offers an inspectable way to vet reading policies before testing them with real students.

Key points

  • Three Wikibooks courses were used. Foundations of Computer Science has 16 chapters, 53 LOs and 131 KCs. General Biology has 20 chapters, 60 LOs and 172 KCs. Introduction to Inorganic Chemistry has 12 chapters, 57 LOs and 177 KCs.
  • Reading is treated as the main intervention, and quizzes serve only as the observation channel for KC-level BKT tracking. The learner's hidden true knowledge is kept separate from the tutor's estimate.
  • Answer choice is score-based rather than free-form LLM generation. This avoids the "competence paradox" of overly capable LLM students and keeps runs reproducible.
  • Adaptive reading led control in every computer science and chemistry cycle. Biology showed a crossover: hidden mastery improved descriptively, but observed accuracy did not.
  • The simulator is not yet calibrated against human data, and it uses fixed heuristic coefficients. Domain differences may partly reflect the quality of the ontology and labeling pipeline.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓Arizona State UniversityTempe · United States50 works · 2,032 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksInformation Systems · 1,437 worksComputer Science Applications · 102 works

🧭 Topics 1Research subjects and languages the work is about.Evaluation & benchmarks · 1,220 works

Cite

Cite as Ryan T. Woo, Anmol Rao, Aryan Keluskar and Yinong Chen. 2026. Evaluating Adaptive Personalization of Educational Readings with Simulated Learners. Proceedings of the 21st Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2026).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.bea-1.63

@inproceedings{woo2026evaluating,
  title = {{Evaluating Adaptive Personalization of Educational Readings with Simulated Learners}},
  author = {Ryan T. Woo, Anmol Rao, Aryan Keluskar and Yinong Chen},
  booktitle = {Proceedings of the 21st Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2026)},
  year = {2026},
  doi = {10.18653/v1/2026.bea-1.63},
  url = {https://doi.org/10.18653/v1/2026.bea-1.63}
}

References

21 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Shivam Bansal and Chaitanya Aggarwal. 2026. textstat: Calculate statistical features from text. Python package, version 0.7.13. Accessed 2026-03-30. find
  2. Jeanne Sternlicht Chall and Edgar Dale. 1995. Readability Revisited: The New Dale–Chall Readability Formula. Brookline Books, Cambridge, MA. find
  3. Albert T. Corbett and John R. Anderson. 1994. Knowledge tracing: Modeling the acquisition of procedural knowledge. User Modeling and User-Adapted Interaction, 4(4):253–278. find
  4. Jennifer G. Cromley and Roger Azevedo. 2007. Testing and refining the direct and inferential mediation model of reading comprehension. Journal of Educational Psychology, 99(2):311–325. find
  5. Jennifer G. Cromley, Lindsey E. Snyder-Hogan, and Ulana A. Luciw-Dubas. 2010. Reading comprehension of scientific text: A domain-specific test of the direct and inferential mediation model of reading comprehension. Journal of Educational Psychology, 102(3):687–700. find
  6. Xinyi Ding and Eric C. Larson. 2021. On the interpretability of deep learning based models for knowledge tracing. Preprint, arXiv:2101.11335. — [arXiv:2101.11335](https://arxiv.org/abs/2101.11335) find
  7. Rudolf Flesch. 1948. A new readability yardstick. Journal of Applied Psychology, 32(3):221–233. find
  8. Mark J. Gierl, Okan Bulut, Qi Guo, and Xinxin Zhang. 2017. Developing, analyzing, and using distractors for multiple-choice tests in education: A comprehensive review. Review of Educational Research, 87(6):1082–1116. find
  9. Xiaoli Huang, Wei Xu, and Ruijia Liu. 2025. Effects of intelligent tutoring systems on educational outcomes: Evidence from a comprehensive analysis. International Journal of Distance Education Technologies, 23(1):1–25. find
  10. Panayiota Kendeou and Edward J. O’Brien. 2014. The knowledge revision components (krec) framework: Processes and mechanisms. In David N. Rapp and Jason L. G. Braasch, editors, Processing Inaccurate Information: Theoretical and Applied Perspectives from Cognitive Science and the Educational Sciences,… find
  11. Walter Kintsch. 1988. The role of knowledge in discourse comprehension: A construction-integration model. Psychological Review, 95(2):163–182. find
  12. Benoît Lemaire, Guy Denhière, Cédrick Bellissens, and Sandra Jhean-Larose. 2006. A computational model for simulating text comprehension. Behavior Research Methods, 38(4):628–637. find
  13. Angélique Létourneau, Marion Deslandes Martineau, Patrick Charland, John Alexander Karran, Jared Boasen, and Pierre-Majorique Léger. 2025. A systematic review of AI-driven intelligent tutoring systems (ITS) in K-12 education. npj Science of Learning, 10(1):29. find
  14. Paul van den Broek and Panayiota Kendeou. 2008. Cognitive processes in comprehension of science texts: The role of co-activation in confronting misconceptions. Applied Cognitive Psychology, 22(3):335–351. find
  15. Kurt VanLehn, Stellan Ohlsson, and Rod Nason. 1994. Applications of simulated students: An exploration. Journal of Artificial Intelligence in Education, 5(2):135–175. find
  16. Wikibooks contributors. 2026a. Foundations of Computer Science. https://en.wikibooks.org/wiki/ Foundations_of_Computer_Science. Accessed 2026-05-13. ↗
  17. Wikibooks contributors. 2026b. General Biology. https://en.wikibooks.org/wiki/General_ Biology. Accessed 2026-05-13. ↗
  18. Wikibooks contributors. 2026c. Introduction to Inorganic Chemistry. https://en.wikibooks.org/ wiki/Introduction_to_Inorganic_Chemistry. Accessed 2026-05-13. ↗
  19. Zhihao Yuan, Yunze Xiao, Ming Li, Weihao Xuan, Richard Tong, Mona T. Diab, and Tom Mitchell. 2026. Towards valid student simulation with large language models. Preprint, arXiv:2601.05473. — [arXiv:2601.05473](https://arxiv.org/abs/2601.05473) find
  20. Tanja Käser and Giora Alexandron. 2024. Simulated learners in educational technology: A systematic literature review and a turing-like test. International Journal of Artificial Intelligence in Education, 34(2):545– 585. find
  21. Panayiota Kendeou. 2024. A theory of knowledge revision: The development of the krec framework. Educational Psychology Review, 36(2):44. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/woo-2026-evaluating-adaptive-personalization-educational-readings/