Optimizing Calculus Learning through Agentic AI: A Process-Oriented Data Management Framework for Undergraduate Mathematics
Year: 2026
Type: conference-paper
Abstract: As undergraduate mathematics education shifts toward hyper-personalized digital environments, the role of data has evolved from a reactive record of performance to a proactive driver of instruction. This study introduces and evaluates a Process-Oriented Data Management (PODM) framework designed to fuel agentic AI tutors in Undergraduate Differential Calculus. Utilizing a mixed-methods, design-based research approach, the study tracked the “micro-learning behaviors” of students at Mapúa University from August 2025 to December 2025. Data were collected from ALEKS and Blackboard clickstream logs, focusing on Process-Oriented Data Markers (PODMs) such as latency and error-correction trajectories. Results indicate that the PODM framework significantly improved student persistence (22% increase) and conceptual mastery$(\mathrm{p}<.05)$. Qualitative analysis revealed a profound shift in students' mathematical argumentation, moving from procedural “Action” to conceptual “Process” understanding. The study concludes that agentic, data-driven scaffolding is essential for fostering metacognitive growth and reducing unproductive frustration in complex STEM subjects.
Authors: Edgar Adina, Rodolfo Santianez, Dante Silva
Institutions: Mapúa University, University of Luzon
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
SDG: Quality education
Open Access status: hybrid
OpenAlex ID: W7204756698