The IRMA Community
Newsletters
Research IRM
Click a keyword to search titles using our InfoSci-OnDemand powered search:
|
The Feedback Loop Architect: A Human-Centered GenAI Platform for Refining Personalized Feedback Ecologies in Science Education
Abstract
Effective feedback is crucial for deep learning, yet providing timely, personalized guidance remains a challenge in education, especially in science. This paper introduces the “Feedback Loop Architect,” a human-centered GenAI platform designed to create personalized feedback ecologies. The platform addresses the need for iterative feedback by simulating and supporting student-platform interactions. A key contribution is our novel methodology for creating a realistic synthetic dataset capturing the dynamics of feedback dialogues. This dataset, informed by established practices like “Sticky Note Feedback,” serves as a testbed for the platform. We demonstrate how the platform simulates learning progressions, visualizes feedback impact, and provides personalized guidance. Results from a simulation of 450 interactions across 50 students show significant improvements in scores (mean increase: 2.69 points) and answer elaboration. This work provides a framework for researchers developing educational AI, offers insights for managers, and explores ethical considerations in such systems.
Related Content
|
Agyabeng N. Yeboah, Leila Goosen.
© 2026.
32 pages.
|
|
Dimitrios Sotiropoulos.
© 2026.
36 pages.
|
|
Dimitrios Sotiropoulos, Christos Xenakis, Apostolis Xenakis, Michail Kalogiannakis.
© 2026.
34 pages.
|
|
Hossein Jamali.
© 2026.
34 pages.
|
|
Nofamataro Zebua, Yan Piter Basman Ziraluo.
© 2026.
26 pages.
|
|
Dipta Gomes, Saikat Baul, Jubayer Ahamed, Dip Nandi.
© 2026.
44 pages.
|
|
María Lucía Barrón-Estrada, Ramón Zatarain-Cabada, Víctor Manuel Bátiz-Beltrán, Aldo Uriarte-Portillo.
© 2026.
40 pages.
|
|
|