Research Seeds Collection

Data Mining

Research

Data mining is a technique for extracting meaningful information from large amounts of diverse and rapidly flowing data.
The types of data we handle are diverse, including data from massive databases, social media data, graph data, and time-series data.
Research themes include: 1) What kind of data will be used? (Research on data), 2) How will it be used? (Research on methodologies and algorithms), and 3) What purpose will it be used for? (Research on applications).
In addition to academic research, we also conduct collaborative research with companies and plan courses on AI and data science.
I'm interested in finding interesting patterns in large amounts of data.

Research achievements related to the aforementioned seed

• NIIMI Ayahiko: Analysis of Factors Affecting the Severity of Traffic Accidents in Hakodate City. IPSJ 181st Database and Data Science & 160th Information Fundamentals and Access Technology Joint Research Conference, IPSJ Research Report IPSJ SIG Technical Report Vol.2025-DBS-181 No.43 Vol.2025-IFAT-160 No.43 2025/9/18
• Ayahiko Niimi: Analysis of Factors Related to Injury Severity in Traffic Accidents Using Machine Learning. International Conference for Internet Technology and Secured Transactions (ICITST2025), 5pages (2025/11/03-5, St Anne's College, Oxford, UK.) (Peer-reviewed)
• NIIMI Ayahiko: Predicting Traffic Accidents with AI – A Collaborative Initiative Between an Information Science University and the Hokkaido Prefectural Police. Fiscal Year 7 2025 2nd Regional ITS Study Group, Hokkaido ITS Promotion Forum, 2025/12/02.

Areas and research topics that can be supported

Data mining is a technique that extracts meaningful information from large amounts of diverse and rapidly flowing data. Therefore, it can be applied to a wide range of fields, not only in research, but also in trend analysis related to corporate production management, real estate valuation, and identification of issues related to preventing traffic accidents in local areas.
This technology isn't simply about having a large amount of data; it requires accumulating data that is appropriate to the nature of the problem. Therefore, it can be helpful in situations like the following:

• I want to predict the number of people using the hotel restaurant based on hotel reservation information.
• I want to estimate the age of a building from its exterior photos.
• I want to find conditions that make traffic accidents more likely to occur by analyzing traffic accident records.
  In addition to academic research, we also engage in collaborative research with companies and plan courses on AI and data science.
• We want to provide employee training on data mining.
• We would like to conduct employee training on data analysis using generative AI.

Experiences and examples of industry-academia collaboration and community contribution, and collaborating companies and organizations.

Hokkaido Prefectural Police Hakodate Regional Headquarters Traffic Division: Development of a traffic accident prediction system (2024-)
Joint research project between Hakodate BASE Architectural Office Co., Ltd. and Yell Co., Ltd.: Research on a real estate valuation system using AI technology (2018)
Data Science Workshop (7 sessions total) (2022-2023) hosted by the Japan Data Science Foundation (JTS).
Data Science Workshop [Basic]: Data Analysis with ChatGPT (3 sessions total) (2024) - Hosted by the Japan Data Science Foundation (JTS)