Towards Building Child-Centered Machine Learning Pipelines: Use Cases from K-12 and Higher-Education
Project Overview
The document explores the application of generative AI in education, specifically through the development of child-centered machine learning (ML) pipelines tailored for K-12 and higher education. It underscores the significance of a human-centered approach to creating ML systems that cater to children's needs and learning experiences. Two illustrative case studies are presented: the first demonstrates the use of predictive modeling to forecast classroom engagement levels, which aids educators in tailoring their teaching strategies, while the second focuses on promoting accessible programming education by utilizing everyday materials for tangible programming activities. Throughout the discussion, the authors highlight critical ethical considerations, including privacy concerns and the necessity of involving children in the design process of these educational technologies. The findings suggest that thoughtfully implemented generative AI can enhance learning experiences, foster engagement, and promote equitable access to educational resources while ensuring that ethical standards are upheld.
Key Applications
Predictive model for classroom engagement levels
Context: K-12 and higher education classrooms, targeting teachers and students
Implementation: Developed baseline architectures, collected data from university students, and created an interactive dashboard for engagement levels.
Outcomes: Provided a tool for teachers to observe and interpret student engagement patterns.
Challenges: Privacy concerns regarding data collection and classroom surveillance.
Tangible programming environment using everyday materials
Context: K-12 education, targeting students and teachers
Implementation: Utilized smartphones and tablets for programming activities with paper programming cards and everyday objects.
Outcomes: Increased access to programming education and enhanced students' understanding of programming concepts.
Challenges: Ensuring seamless integration of everyday objects in programming activities.
Implementation Barriers
Ethical Barrier
Concerns regarding privacy and potential classroom surveillance with camera usage.
Proposed Solutions: Developing systems that do not require data collection from children and considering on-device processing.
Project Team
Alpay Sabuncuoglu
Researcher
Ceylan Besevli
Researcher
T. Metin Sezgin
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Alpay Sabuncuoglu, Ceylan Besevli, T. Metin Sezgin
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