Enhancing Online Learning Efficiency Through Heterogeneous Resource Integration with a Multi-Agent RAG System

TL;DR

This work tackles the challenge of efficient online learning in the presence of diverse online resources by proposing a Multi-Agent Retrieval-Augmented Generation (RAG) system. It employs a Manager Agent and four specialized resource agents (YouTube, GitHub, documentation sites, and general web search) powered by GPT-4o, with semantic context stored in ChromaDB and a Streamlit UI for user interaction. The approach aims to automate retrieval and synthesis to reduce manual cross-source effort and improve learning efficiency, and it receives preliminary validation through a TAM-based user study with 15 participants showing strong usability and moderate-high utility. The results suggest practical impact in aiding targeted, real-time information integration, with future work focusing on dynamic agent selection, broader domain coverage, and user-personalization.

Abstract

Efficient online learning requires seamless access to diverse resources such as videos, code repositories, documentation, and general web content. This poster paper introduces early-stage work on a Multi-Agent Retrieval-Augmented Generation (RAG) System designed to enhance learning efficiency by integrating these heterogeneous resources. Using specialized agents tailored for specific resource types (e.g., YouTube tutorials, GitHub repositories, documentation websites, and search engines), the system automates the retrieval and synthesis of relevant information. By streamlining the process of finding and combining knowledge, this approach reduces manual effort and enhances the learning experience. A preliminary user study confirmed the system's strong usability and moderate-high utility, demonstrating its potential to improve the efficiency of knowledge acquisition.

Figures (3)

• Figure 1: Working Flow of the Multi-Agent RAG System
• Figure 2: Analysis of the user study using TAM
• Figure 3: User Interface of the proposed system.