Related Experiment Videos
Development and effectiveness verification of AI education data sets based on constructivist learning principles for
Seul-Ki Kim1, Tae-Young Kim2, Kwihoon Kim3
1Department of Computer education, Korea National University of Education, Chungju, Chungbuk, Republic of Korea. tmfrlska85@gmail.com.
Scientific Reports
|March 29, 2025
Summary
This study developed new AI education datasets grounded in constructivism, enhancing students' AI literacy by connecting learning to real-world experiences. These datasets offer a valuable alternative to traditional materials for effective AI education.
Area of Science:
- Artificial Intelligence Education
- Computer Science Education
- Educational Technology
Background:
- Traditional AI education often lacks contextual relevance for students.
- Existing datasets may not adequately support constructivist learning principles.
- There is a need for AI educational resources that foster deep AI literacy.
Purpose of the Study:
- To develop and evaluate constructivist-oriented AI education datasets.
- To enhance students' AI literacy through real-world problem-solving.
- To create sustainable and accessible AI educational resources.
Main Methods:
- Reconstructed the machine learning dataset development cycle.
- Developed and refined AI datasets through expert panel interviews.
- Deployed datasets on educational platforms and analyzed usage metrics.
- Conducted comparative analysis of AI literacy impact.
Main Results:
- Developed four novel AI education datasets suitable for replacing conventional ones like the Iris dataset.
- Confirmed high accessibility and utility on major Korean AI education platforms.
- Demonstrated effectiveness in enhancing AI literacy through practical application.
Conclusions:
- Constructivist AI datasets effectively connect prior knowledge with real-world experiences.
- These datasets deepen understanding of AI model learning processes.
- They provide authentic, data-driven computing experiences crucial for AI literacy.
