A new undergraduate research student has joined our laboratory
About
Turning Diverse Data into Trusted Knowledge
We develop data engineering methods to make diverse real-world data reliable, discover meaningful patterns, and explain their variations.
Our research combines database systems, data mining, statistical analysis, machine learning, and trustworthy data management to transform heterogeneous real-world data into knowledge that can be understood and reliably used.
Questions that Connect Our Research
01
Can we trust the data?
Data Quality
How can incomplete, inconsistent, or heterogeneous real-world data be made reliable enough for subsequent analysis?
02
What structures and patterns exist in the data?
Pattern Discovery
How can we discover meaningful and recurring structures from large and complex datasets?
03
Why do these patterns differ?
Variation Explanation
How can we explain why different patterns, behaviors, or branching processes emerge?
04
What data and evidence support the result?
Trustworthy Data Management
How can analytical and AI-generated results be traced back to the data and evidence on which they are based?
Affiliations: Department of Humanities Data Engineering (Undergraduate), Department of Transdisciplinary Engineering and Department of Computer Science (Master’s) and Computer Science Program (Doctoral) of Ochanomizu University.
Representative Research
From Medical Data to Variations—and Their Explanations
Electronic medical records from multiple healthcare institutions contain examinations, medications, procedures, and other medical orders recorded as temporal sequences. Although many treatment processes are shared, variations occur across patients and healthcare institutions.
Our research analyzes these data through the following process:
01 Discover frequent treatment patterns
02 Compare differences across institutions and patients
03 Identify where treatment processes branch
04 Explain factors associated with those branches
Our goal is not merely to show that different patterns exist. We aim to understand why those differences occur and ultimately what evidence supports their explanation.
Beyond Healthcare
The Data May Change, but The Data-Engineering Questions Remain
Cultural & Multimodal Data
We study music, images, audio, movies and video, dance, performing arts, and other forms of cultural and multimodal data. From a data-engineering perspective, we investigate similarities, recurring structures, temporal patterns, and differences among works, behaviors, and expressions.
Industrial & Societal Data
We study data that are distributed across multiple organizations and systems. Our research addresses problems such as data integration, data integrity, tamper detection, provenance, and traceability.
Fundamental Data Engineering
Our research is not limited to a particular application domain. We investigate fundamental problems in data quality, databases, information retrieval, data mining, information storage, data privacy, and trustworthy data management. Even when the type of data changes, the fundamental data-engineering questions we ask remain the same.
News
Recent research presentations, awards, and laboratory activities.
We have significantly expanded the information on our research, laboratory activities, and resources for prospective students
Sugitani, a second-year Master’s student in our laboratory, presented research on factor estimation in treatment patterns at IEEE ISMICT 2026
The Yume Navi article and mini-lecture video towards high school students, titled “Understanding the Treatment Process through the Use of Medical Data,” (in Japanese) have been published
Two new master students joined our lab