Integrating Cognitive AI with Generative Models for Enhanced Question Answering in Skill-based Learning
Project Overview
The document explores the integration of Cognitive AI and Generative AI in education, specifically focusing on enhancing question-answering in skill-based online learning environments. It highlights the importance of improving skill understanding and explanation within educational contexts. To address this, a framework is proposed that utilizes the TMK (Task-Method-Knowledge) model to better represent skills. The AI system, named Ivy, is designed to generate reasoned explanations in response to learners' questions by leveraging advanced techniques such as Large Language Models, Chain-of-Thought reasoning, and Iterative Refinement. The findings indicate that such applications of generative AI can significantly enhance the learning experience by providing personalized and context-aware feedback, thereby fostering deeper comprehension and skill acquisition among learners. Overall, the document underscores the potential of generative AI to transform educational practices and improve outcomes in online learning settings.
Key Applications
Ivy - Interactive Video system
Context: Online learning environments, particularly for adult learners seeking to reskill or upskill
Implementation: The TMK model is used to represent skills and guide the question-answering process, leveraging Generative AI methods for response generation.
Outcomes: Provides contextually rich and reasoned explanations that enhance learner engagement and understanding of skills.
Challenges: Generative AI methods may lack true understanding of nuanced skills and can produce inconsistent or inaccurate responses.
Implementation Barriers
Technical Barrier
Generative AI methods may not exhibit true understanding and can produce inconsistent outputs.
Proposed Solutions: Integrating Cognitive AI methods to enhance understanding and improve response accuracy.
Scalability Barrier
Manual creation of TMK models for new educational content is time-consuming and requires domain expertise.
Proposed Solutions: Automation of the TMK model creation process to streamline setup for diverse learning domains.
Project Team
Rochan H. Madhusudhana
Researcher
Rahul K. Dass
Researcher
Jeanette Luu
Researcher
Ashok K. Goel
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Rochan H. Madhusudhana, Rahul K. Dass, Jeanette Luu, Ashok K. Goel
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