Designing AI-Enabled Games to Support Social-Emotional Learning for Children with Autism Spectrum Disorders
Project Overview
The document discusses the innovative application of generative AI in an educational context, specifically through the design of an AI-enabled mobile game tailored for children with Autism Spectrum Disorder (ASD) to enhance their social-emotional learning. This game leverages AI to generate personalized stories that aid children in recognizing and understanding emotions via interactive scenarios that resonate with their experiences. The development process involved collaboration with experts in the field to ensure the game's effectiveness, and a fine-tuned AI model was employed to create a diverse dataset of social stories that accurately reflect various emotions. Key insights from the game’s design highlight the importance of adaptability and engagement by using familiar contexts, alongside providing multimodal feedback to maximize learning outcomes for children with ASD. Overall, the use of generative AI in this educational setting demonstrates promising potential for supporting emotional understanding and social skills development among children with special needs.
Key Applications
AI-enabled mobile game for social-emotional learning
Context: Educational intervention for children with Autism Spectrum Disorder (ASD) and their caregivers
Implementation: Co-design process with experts, using a fine-tuned AI model (GPT-4) to generate personalized stories based on emotional scenarios
Outcomes: Improved emotional understanding, facial expression recognition, and engagement in social-emotional learning for ASD children
Challenges: Financial constraints for caregivers, limited access to resources, and the challenge of personalizing stories for diverse needs
Implementation Barriers
Access Barrier
Caregivers face financial constraints and limited access to services and resources for interventions.
Proposed Solutions: Developing tools to promote convenience and affordability for families with autistic children.
Project Team
Yue Lyu
Researcher
Pengcheng An
Researcher
Huan Zhang
Researcher
Keiko Katsuragawa
Researcher
Jian Zhao
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Yue Lyu, Pengcheng An, Huan Zhang, Keiko Katsuragawa, Jian Zhao
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