AI Education For K12
2020. Xiaoyu Wan, Xiaofei Zhou, Zaiqiao Ye, Chase K. Mortensen, and Zhen Bai. "SmileyCluster: supporting accessible machine learning in K-12 scientific discovery". In Proceedings of the 2020 Conference on Interaction Design and Children (IDC 2020).
2019. Zhen Bai, Zaiqiao Ye, Xiaoyu Wan, Zhaoxiong Ding. "FaceOverlay: Supporting Learning of Cluster Analysis for Scientific Discovery." AIED 2019: International Conference on Artificial Intelligence in Education. Accepted to Workshop on “K12 AI education”. Chicago, IL, United States, June 25, 2019
*In this project team, I worked as a master thesis intern. I lead the design and development of the learning system, including the data visualization of the ML narratives. I also participated in the organization of the learning workshop and analyzed the data. This project also serves as part of my master thesis.
Introduction
There is an increasing need to prepare young learners to be Artificial Intelligence (AI) capable for the future workforce and everyday life. Machine Learning (ML), as an integral subfield of AI, has become the new engine that revolutionizes practices of knowledge discovery. Making ML experience accessible to young learners, however, remains challenging due to its high demand for mathematical and computational skills. This research focuses on designing novel learning environments that help demystify ML technologies for K-12 students, and also investigating new opportunities for maximizing ML accessibility through integration with scientific discovery in STEM education. We developed SmileyCluster - a hands-on and collaborative learning environment that utilizes glyph-based data visualization and superposition comparative visualization to assist learning an entry-level ML technology, namely k-means clustering. Findings from an initial case study with high school students in a pre-college summer program show that SmileyCluster leads to positive change in learning ML concepts, methods and sense-making of patterns. The findings of this study also shed light on understanding ML as a data-enabled approach to support evidence-based scientific discovery in K-12 STEM education.
Visualization of Machine Learning Data Points
(Designed together with Dr. Bai Zhen)
Learning System Design
(Designed together with Dr. Bai Zhen, developed with Chase K. Mortensen)
Learning Workshop
(Organized and analyzed together with Xiaoyu Wan, Xiaofei Zhou and Dr. Bai Zhen)
Participants
We recruited eight participants through the program coordinator of the on-campus pre-college summer program with a total of 12 students enrolled. There were 5 female and 3 male students, between 15-17 years old, with 5 domestic and 3 international students all with sufficient language skills for the class.
Procedure
The study took place in an on-campus computer lab and lasted about 2.5 hours, facilitated by one course instructor and four researchers. When the students showed up in the classroom,
one researcher arranged for students who did not consent to be involved in the study to sit separately with participants of the study. Then the participants sat in pairs of their choice
in front of a computer. The study procedure was (1) 25 minutes class instruction about AI and general difference between supervised and unsupervised learning, without revealing any content about clustering; (2) 15-min pre-study questionnaire about students’ background in machine learning and cluster analysis; (3) 40-min interaction with the system; (4) 15-min post-study questionnaire and (5) 30-min focus group interview. All students in the class took part in the same learning activities, except we did not collect any observation data from
students who did not take part in the study.
