Naruhiko Shiratori
  • Institutional Research
  • Dropout Prediction
  • Learning Analytics
  • Data Science Education

Predict, act, verify. 

Predicting dropout alone does not help students. I connect prediction to support, measure the effect, and feed it back into the next prediction, and I build that cycle into how a university works.

Naruhiko Shiratori

白鳥 成彦
  • Professor, Faculty of Liberal Arts and Sciences, Tokyo City University
  • Center for Educational Assessment and Institutional Research
  • Special Assistant to the President
Publications on researchmapContact
Illustration of Naruhiko Shiratori holding a laptop and pointing at a rising line chart, with a database, students, and a university building at his feet

What I am working on

  • Research

    Building a cyclic dropout-prevention model

    Linking pre-admission data with daily attendance and grades to predict risk, connect it to support, evaluate the intervention, and return the result to the next prediction. Funded by JSPS KAKENHI (FY2026–2030).

  • Practice

    Running an institutional research data platform

    Bringing enrollment, grade, attendance, and course data that used to live in separate offices onto Microsoft Fabric, refreshed automatically and available to staff through Power BI. In operation at Tokyo City University since FY2025.

  • Education

    Teaching students and staff to work with data

    First-year data science and information literacy courses, plus data workshops for faculty and staff. Research findings go back into teaching and training.

Recent publications

Educational technologySociology of educationIntelligent informaticsIR

All 36 papers and 33 other items

Institutional research in practice

  • Institutional research data platform (Microsoft Fabric)

    Four data sources, enrollment, grades, attendance, and courses, managed in raw, refined, and curated layers, refreshed monthly and daily, and served to Power BI and apps. It underpins dropout reduction, learning-outcome visualization, and assessment.

    • Microsoft Fabric
    • Power BI
    • Medallion architecture

  • Early warning and dropout reduction

    Detecting changes in attendance and grades early in the semester and routing students who need help to student-support staff. Research prediction models implemented in day-to-day operation.

    • Dropout prediction
    • Attendance data
    • Student support

  • Learning outcomes and assessment

    Bottom-up indicators of diploma-policy attainment, a university data book, and assessment days that give academic management an evidence base.

    • Learning outcomes
    • Assessment
    • Data book

  • Data workshops for faculty and staff

    IR workshops and onboarding sessions where staff look at data together and discuss it. I also give talks and training for other universities and organizations.

    • Faculty development
    • Workshops
    • Talks

Career

2024 — present
Professor, Tokyo City UniversityFaculty of Liberal Arts and Sciences
2021 — 2024
教授, Kaetsu University経営経済学部
2019 — 2021
教授, Kaetsu Universityビジネス創造学部
2013 — 2019
准教授, Kaetsu Universityビジネス創造学部
2009 — 2013
専任講師, Kaetsu University

Degrees

Doctor of Engineering
Tokyo Institute of Technology
Master of Media and Governance
Keio University

Education

2017 — 2023
Tokyo Institute of TechnologySchool of Environment and Society, Department of Social and Human Sciences
2001 — 2009
Keio UniversityGraduate School of Media and Governance, Major in Media and Governance
1994 — 1999
Chiba UniversityFaculty of Engineering, 工業意匠学科

Consulting, talks, and training

I am happy to talk about university data platforms, connecting dropout prediction to student support, setting up an IR function, and designing data science education.