Center for Mathematical Modeling and Data Science,
I work on artificial intelligence (AI). My team and I specialize in a field called multi-agent systems among diverse fields related to AI. Real society works based on countless decisions, which are made by many people who exchange information with each other. Our research focuses on the cooperation and competition among these people. Specifically, we explore how to create a mechanism to make such cooperation/competition work well.
For example, let's say you want to predict inflation index trends. Predictions using machine learning depend on quantitative data such as past price fluctuations. However, non-quantitative data such as current fiscal policies or the personality of a state leader cannot be used as inputs by today's machine learning solutions.
My research team strives to improve the prediction accuracy of human-AI interactions with the assumption that collective intelligence, where people and a machine-learning computer communicate and complement each other, achieves better decision making. We are also exploring how to incorporate people's opinions to make quality predictions, fully leveraging the diversity of individuals' opinions.
At MMDS, I teach many undergraduate courses from all Schools of Osaka University. For example, the course "Information and Society," which was offered in the first semester this year, was a liberal arts course geared towards first- and second-year undergraduate students, not only from the information field but also from fields such as economics, law, and foreign languages. Although I personally don't like to set a border between arts and sciences, I would like students in majors that don't have an opportunity to write programs to acquire the ability to communicate well with people who develop information systems as this is necessary when they start working in the real world. This course is designed to help students recognize what kind of information technology is being used in various parts of the world and its connection to society.
In the course, four lectures are provided in the form of workshops. For example, in one session students discuss the theme of “how online lectures can be improved?" in pairs and then in a team of three to further expand their ideas. Then they give presentations on their conclusions.
No, MMDS also provides education for students from other universities as well as for working professionals. As part of such an education framework, I offer e-Learning courses for students and working professionals through Human Resource Association of Mathematics (HRAM) activities.
Of course, working professionals and even current undergraduate students can take this e-Learning course when lectures are unavailable to study topics that interest them at their university or school/faculty/department. Students at Osaka University can further deepen their knowledge through HRAM's e-Learning in addition to on-campus classes offered by MMDS. Anyone can participate in our education using the method of learning that meets their individual situation.
I also offer a wide range of courses, from basic to practical, where synergistic effects can be expected. For example, by reflecting on what I learned during lectures for working professionals, I improved the quality of lectures offered to first-year undergraduates. I think it's beneficial to have a variety of things and people mixed together, rather than compartmentalizing as it results in a lively atmosphere. In the future, I would like to make the connections or links between different courses clearer. This will help highlight the relationships among courses, allowing participants to choose subsequent courses that will deepen their knowledge.
The D-DRIVE (Doctoral program for Data-Related InnoVation Expert), a data-related human resource development program for which MMDS serves as one of the co-lead organizations, holds an "interactive matching" event twice a year to arrange internship opportunities between students and companies.
Recently, due to the COVID-19 pandemic, these events have been held in a hybrid manner, using both face-to-face and online sessions. In the event, participating students and company representatives meet. The students first give a one-minute presentation on what kind of research they are doing and what topics they are interested in. Next, each company explains its internship program and what business it conducts. Then, students are free to visit company booths where each company's representatives are waiting to talk to them.
So far, many graduate students from Osaka University as well as other universities across Japan have participated in the interactive matching events, which have resulted in successful internship programs. Companies representing diverse sectors from informatics to industry have participated in the events.
For students, the matching events provide opportunities to efficiently become familiar with a variety of internship programs, which previously could only be found by searching individually. Additionally, students may encounter companies whose businesses are aligned with their interests but were not on their to-visit lists.
Internship programs also provide a valuable opportunity to come into contact with "real-world data." It is said that data is everything in data sciences, but most data handled in universities is prepared for practice purposes and is not produced in real businesses. For example, a real factory is a gold mine of data. The massive amount of accumulated data is not well utilized. Although production efficiency can be improved by analyzing such data, factories lack the labor resources to do so.
In such a context, it is advantageous for companies to meet talented students in internship programs and see the latest data science methods that they can handle. Data scientists are still few and far between. Through these matching events, we can help resolve the mismatch between students’ and companies’ needs by providing a scheme to know more about companies’ businesses, which cannot be gained by merely visiting companies’ websites. Actually, in our questionnaire surveys, about 90% of participants answered that they were satisfied and happy with the event.
Interview as of January 2022
*Interview and photography were conducted with countermeasures against COVID-19.