Date
Publisher
arXiv
Learning to learn is becoming a science, driven by the convergence of
knowledge tracing, signal processing, and generative AI to model student
learning states and optimize education. We propose CoTutor, an AI-driven model
that enhances Bayesian Knowledge Tracing with signal processing techniques to
improve student progress modeling and deliver adaptive feedback and strategies.
Deployed as an AI copilot, CoTutor combines generative AI with adaptive
learning technology. In university trials, it has demonstrated measurable
improvements in learning outcomes while outperforming conventional educational
tools. Our results highlight its potential for AI-driven personalization,
scalability, and future opportunities for advancing privacy and ethical
considerations in educational technology. Inspired by Richard Hamming's vision
of computer-aided 'learning to learn,' CoTutor applies convex optimization and
signal processing to automate and scale up learning analytics, while reserving
pedagogical judgment for humans, ensuring AI facilitates the process of
knowledge tracing while enabling learners to uncover new insights.
Study at a glance
- Question: Can combining knowledge tracing, signal processing, and AI copilots improve student learning and engagement?
- Setting and data: University AI course, one semester, 203 students; public EdNet KT1 dataset.
- Method: Randomized controlled experiment across five groups and benchmark evaluation on public dataset.
Why it matters
Generative AI systems often rely on end-to-end prompting without tracking students' learning trajectories, limiting their ability to dynamically adapt instruction to individual needs in real time.
Key findings
- Students using the CoTutor system achieved the highest average assignment score (73.2%), outperforming baseline models and the traditional control group (68.4%).
- CoTutor attained higher performance prediction accuracy (67.4%) and post-course survey satisfaction scores (72.8%) compared to all baseline models and the control.
- On the public EdNet dataset, CoTutor achieved higher accuracy (84.3%) and lower prediction error than traditional and deep knowledge tracing baselines.
Summary drafted with AI from the full paper and may contain errors. See the full paper for details.
What is the application?
Who age?
Why use AI?
Study design
