Title: SortingHat: Transforming Operating Systems Pedagogy Through Personalized AI Assistance

Abstract

Computer science curricula frequently identify Operating Systems (OS) courses as particularly demanding, largely due to the intricate nature of internal system architectures and the wide variety of execution environments involved. Conventional instructional approaches often struggle to accommodate the varying academic backgrounds, paces of learning, and specific practical requirements of individual students. In response to these educational hurdles, this paper introduces SortingHat, a specialized digital teaching assistant designed to personalize OS learning.

SortingHat leverages sophisticated artificial intelligence, specifically combining a Retrieval-Augmented Generation (RAG) framework with Multi-Agent Reinforcement Learning (MARL), to offer educational support that is both scalable and adaptive. The platform features an immersive 3D digital human interface driven by Large Language Models (LLMs), enabling it to provide guidance that is context-aware, empathetic, and highly individualized.

To enhance learning outcomes, the system analyzes each student’s historical performance and academic data to generate customized exercises. This approach targets knowledge gaps by reinforcing weak areas while simultaneously presenting advanced concepts to proficient learners. Furthermore, SortingHat employs a rigorous evaluation pipeline to ensure that the assessment of student submissions is consistent, unbiased, and fair. This automated grading system is complemented by personalized, actionable feedback aimed at fostering improvement. By synthesizing adaptive content generation, personalized mentorship, and automated assessment, SortingHat reimagines OS education as an engaging, immersive, and scalable experience.

Source: arXiv Generated at: 2026-06-02 00:00:00 UTC