Transforming E-learning: a chatbot-based approach to personalized course recommendations

Abstract

This project presents an innovative chatbot system designed to deliver personalized course suggestions for e-learning platforms. It addresses common issues in lifelong learning, including fragmented content, a lack of personalization, and the shortcomings of current chatbots. At the core of this online web application, called "Journey," lies a combination of multiple recommendation algorithms, including content-based filtering, knowledge-based filtering, user input, matrix factorization, and graph clustering, within an agent based chatbot design. The system utilizes Neo4j to visualize course relationships, manages user data through Firebase, and enhances semantic search capabilities using Pinecone. The implementation includes specialized agents—Guard, Classification, General, Detail, and Recommendation—that effectively handle various types of inquiries while facilitating fluid conversations. While challenges related to data access and natural language comprehension remain, this system represents a notable advance in personalized e-learning, establishing a strong foundation for future developments in adaptive learning technologies.