Abstract:

Amid rapid societal change, there is an increasing need to cultivate talent capable of generating new research and educational practices from a global perspective and creating new social value. We argue that developing students’ ability to address societal challenges by applying university-acquired knowledge and contemporary technologies is critically important, particularly for early-year undergraduates. This study proposes an educational methodology that integrates AI into undergraduate education using Google’s machine-learning platform, Teachable Machine, to support research-oriented learning. Within this framework, lower-year undergraduates are guided to identify social problems, formulate research questions, and explore solutions through AI-assisted inquiry, structured around the three dimensions of “what to learn,” “what to solve,” and “how to solve it.”

The proposed approach was implemented with a third-year undergraduate student who, over a seven-month period, developed an original research proposal and demonstrated independent problem-solving ability. As a result, the student successfully presented the research at an international conference in November 2025. These results indicate that combining accessible machine learning tools can effectively support research engagement and inquiry-based learning among early-year undergraduate students.

Keywords:

Educational, AI, undergraduate student, machine learning tools.