Explanation as Question Answering based on Design Knowledge
Project Overview
The document explores the implementation of generative AI in education through a question-answering agent named AskJill, which is integrated into an interactive learning environment called VERA. AskJill leverages the User Guide of VERA to deliver explanations about its domain, functionality, and operation, thereby enhancing the accessibility and quality of online education. VERA facilitates virtual experiments, especially in the field of ecology, and promotes self-regulated learning through model construction and simulation. Additionally, the paper addresses the challenges associated with AI-generated explanations and underscores the significance of user feedback in refining AI systems. Overall, the integration of generative AI in educational settings, exemplified by AskJill within VERA, demonstrates its potential to support effective learning experiences while also highlighting areas for improvement in AI-driven educational tools.
Key Applications
AskJill, a question-answering agent for VERA
Context: Online education, particularly for learners in ecology and biology, including students and citizen scientists.
Implementation: AskJill is integrated into the VERA platform, using a REST API to process user questions and retrieve answers from the VERA User Guide.
Outcomes: Provides immediate, context-relevant answers to user questions, enhancing comprehension of VERA's domain and functionality.
Challenges: Limited ability to answer questions outside its training data and the need for continuous updates to the User Guide based on user feedback.
Implementation Barriers
Technical barrier
AskJill may not answer all user queries accurately, particularly if they fall outside its training or if the User Guide lacks certain information.
Proposed Solutions: Collect user feedback to identify gaps in knowledge and update the User Guide accordingly.
User engagement barrier
Users may not read the User Guide in detail, limiting their understanding of the AI system.
Proposed Solutions: Implement an interactive User Guide format that engages users more effectively and prompts them to ask questions.
Project Team
Ashok Goel
Researcher
Vrinda Nandan
Researcher
Eric Gregori
Researcher
Sungeun An
Researcher
Spencer Rugaber
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Ashok Goel, Vrinda Nandan, Eric Gregori, Sungeun An, Spencer Rugaber
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