Knowledge Navigated and Organized Together(KNOT): An AI-Integrated System Design for Enhanced Collaborative Learning in Higher Education

Abstract

Collaborative learning in higher education is often hindered by the cognitive overhead of administrative tasks and fragmented, passive digital tools. This research introduces KNOT, a proactive AI-driven system designed to serve as an integrated facilitator, administrator, and tutor for student study groups. Utilizing a multi-phase human-centered design approach, we conducted ethnographic observations (5 sessions, 15 hours), surveys (N=54), and semi-structured interviews (N=6) with undergraduate STEM students. The research identified four primary challenges: administrative friction, conceptual deadlocks, lack of group synchronicity, and dependence on scarce physical resources. In response, we developed and prototyped the KNOT ecosystem, featuring automated scheduling, AI-powered Socratic tutoring, interactive content synthesis, and personalized learning paths. User testing (N=5) validated that KNOT successfully offloads logistical burdens, with participants reporting improvement in their ability to focus their cognitive energy on learning. The AI tutor effectively guided groups through conceptual roadblocks without replacing human critical thinking. Furthermore, the integration of interactive study aids and gamified features directly mitigated focus loss, a primary challenge identified by 48% of surveyed students. KNOT demonstrates how an adaptive AI can seamlessly organize, enhance, and support collaborative learning.

Keywords: Computer-supported Collaborative Learning, AI Agents, Educational Technology, Human-AI Interaction, Systems Design, Product Design, UX Design

DOI: 10.54941/ahfe1007765