Skip to main content Skip to navigation

Automated Personalized Feedback Improves Learning Gains in an Intelligent Tutoring System

Project Overview

The document explores the application of generative AI in education, focusing on its role in delivering automated, personalized feedback through intelligent tutoring systems (ITS). A prime example is the Korbit platform, which utilizes machine learning and natural language processing to generate customized hints and explanations for students. This adaptive system tailors its responses to meet individual learning needs, offering diverse feedback formats such as mathematical hints and Wikipedia-based explanations. Research findings demonstrate that such personalized feedback not only enhances student engagement but also significantly improves learning outcomes, highlighting the transformative potential of generative AI in educational settings. The emphasis on individualized support showcases how technology can effectively address diverse learning challenges and foster a more effective educational experience.

Key Applications

Korbit Intelligent Tutoring System (ITS)

Context: Large-scale online learning environment for topics related to data science, machine learning, and artificial intelligence, targeting students enrolled in these courses.

Implementation: The ITS uses machine learning and natural language processing techniques to generate personalized hints and explanations based on student interactions and performance.

Outcomes: Considerable improvement in student learning outcomes and subjective evaluation of the feedback, with deep personalization models showing the highest student learning gains.

Challenges: The complexity of simulating diverse student-tutor interactions and generating contextually appropriate feedback.

Implementation Barriers

Technical Barrier

The challenge of managing the multitude of scenarios in student-tutor interactions within an ITS.

Proposed Solutions: Utilizing automated, data-driven feedback generation methods to enhance the scalability and adaptability of the feedback mechanism.

Project Team

Ekaterina Kochmar

Researcher

Dung Do Vu

Researcher

Robert Belfer

Researcher

Varun Gupta

Researcher

Iulian Vlad Serban

Researcher

Joelle Pineau

Researcher

Contact Information

For information about the paper, please contact the authors.

Authors: Ekaterina Kochmar, Dung Do Vu, Robert Belfer, Varun Gupta, Iulian Vlad Serban, Joelle Pineau

Source Publication: View Original PaperLink opens in a new window

Project Contact: Dr. Jianhua Yang

LLM Model Version: gpt-4o-mini-2024-07-18

Analysis Provider: Openai

Let us know you agree to cookies