The ML-Machine Toolkit: Empowering Teachers and Education Professionals to Explore Embodied Approaches to Teaching Machine Learning

Most HCI studies on teaching K-12 students about machine learning (ML) through embodied interaction approaches are based on design and evaluation of one-off prototypes and are not sustained in schools after the studies. In addition, the tools are seldom theoretically positioned, which makes the overall research effort largely technology-driven. This work presents an HCI toolkit, ML-Machine, for supporting teachers and education professionals in developing and conducting embodied educational activities with ML. It encapsulates theory and intermediate-level knowledge from previous HCI research in three design principles - \textit{enacting ML practices, using ML as a design material, and embodied exploration of ML} - to make them readily available to be integrated into educational contexts and practices. We evaluate the toolkit through a case study with a teacher, library employees, and content developers. Based on this, we discuss how toolkits can develop HCI research efforts on teaching digital emerging technologies in K-12 education.