Topics that shape the future
- Closer look at research results -

The field I research, "cognitive science," is the study of how humans perceive things and incorporate them into their knowledge. People from various research fields come together, primarily aiming to unravel the mysteries of "knowledge," that is, intelligence and learning. During my university years, I majored in cognitive science, and throughout my studies, I have continuously researched questions such as "how do people learn?" and "what kind of support is possible for better learning?"
The field of cognitive science encompasses not only school education but also all aspects of human intellectual activity in daily life and various situations. However, since the 1990s, there has been a surge in efforts to apply its findings to educational settings. "Learning science" emerged from this trend. Learning science places importance on the use of diverse technologies, including artificial intelligence (generative AI). Moving beyond the traditional framework of school education, the main focus of current learning science is to build better forms of learning by utilizing cutting-edge technologies such as generative AI.

Learning science is based on learning theories such as "how people learn" and "when people learn best." By combining decades of accumulated research with new information technologies, it becomes possible to design richer learning and educational environments. Through learning science, we hope to build goal-setting and learner-centered learning environments, and as a result, contribute to positive change in society.

For example, in the medical field, it is common practice to collect numerous cases using different methods and scientifically compare and analyze them in order to measure the effectiveness of drugs and treatments. On the other hand, in the field of education, such quantitative comparisons are difficult, and until now, there has been a tendency to rely on the empirical rules of excellent educators. Education of the future will require that such experiences not be treated as knowledge that remains within the individual, but rather recorded and analyzed as data and made into a form that can be shared with the next generation of educators. My research goal is to clarify the ideal form of goal-setting and learner-centered education through this scientific approach, and to return the results to practice.

As stated on their official website, the Japan Kanji Aptitude Testing Association operates with three main pillars: "Public Awareness and Educational Support Activities," "Research and Study Activities," and "Japanese Language Proficiency Development Activities," aiming to improve overall Japanese language proficiency, including expressive ability and vocabulary. As part of these activities, they approached us to explore the possibility of developing new service formats by applying insights from learning science, which led to this project.
Furthermore, this time, the use of generative AI is also included, and the "words" handled by the Japan Kanji Aptitude Testing Association and the "linguistic information" handled by generative AI are in the same domain. I think this was also a time when we wanted to explore new ways of learning that support learners' Japanese language abilities, without being bound by traditional teaching methods and flexibly incorporating advances in generative AI technology. In the field of cognitive science, including learning science, there is a long history of research into the use of AI, so I feel that various elements have come together at this time.

The primary reason for conducting collaborative research is that I believe companies and organizations, by providing educational services to various sectors, contribute to education in society. By working together with these entities to explore better ways of education, I hope that this will ultimately lead to improvements in society as a whole. In reality, many companies have not yet fully incorporated a learning science perspective, and in many cases, services are being developed using conventional educational designs. In such cases, there may be limitations to the growth of the target audience. That is why I see great value in engaging with the actual sites where services are being provided through collaborative research.
In the field of learning science, "social contribution" and "social implementation" have become increasingly important over the past decade or so, and the concept of "implementational research," which focuses on practical development, has gained attention. For me personally, one of my major research themes is how to spread the "future of education and learning" that we have seen through our research and practice to society as a whole.

Before and after this learning experience, students were asked to write down on a worksheet what they considered to be the most important points when writing a self-introduction essay. By analyzing their responses, we were able to evaluate the system by checking how many students came to understand, for example, how much more students realized the importance of clearly distinguishing between facts and opinions after going through this learning activity, what changes were observed in their actual writing, and whether students were able to engage in deeper dialogue with each other based on the AI-generated answers.
However, these kinds of writing skills cannot be acquired in a short period of time. Therefore, in the future, we need to carefully examine whether students can properly utilize generative AI in continuous learning, such as in junior high school Japanese language classes, and whether this technology will become firmly established as their own skill, or whether the effect will be limited to a temporary one immediately after using the app.

In public junior high schools in Sagamihara City, teachers are using pre-prepared prompts to allow students to answer questions from a generated AI and deepen their understanding of the material through dialogue with other students based on their answers. For example, they are experimenting with improving the quality of endurance running by leveraging the strengths of both humans and AI, with the help of classmates and the generated AI. This is a challenge to see how AI can be used in learning in various fields, not just core subjects, but also art and physical education.
What's essential in these initiatives is not simply accepting the AI's answers at face value, but creating opportunities for students to think for themselves, such as discussing the answers with their classmates. If generative AI is introduced into learning without much thought, students tend to become passive, thinking, "If I ask the AI, I'll get the answer." However, AI is merely an "additional" element; the core of learning should be dialogue with others and the ability to think for oneself. We want to continue to nurture this kind of learning in the future.