In recent years, there are growing expectations around collecting and utilizing large amounts of personal data from IoT devices, smartphones, car navigation systems, tablets, smart meters, and other sources. For example, researchers are studying crowd sensing applications that collect location and conversation data to analyze popular tourist destinations, road traffic information, and products of interest. Additionally, by collecting attribute data such as age, marital status, income, education level, and product satisfaction ratings, it becomes possible to estimate underlying data distributions, perform correlation analysis, and clearly identify target customers. However, there are simultaneous concerns about privacy violations. For instance, there have been actual cases of burglaries and stalking incidents using publicly available location data, and there are risks that conversation data could be used to infer people's lifestyle patterns.

Against this backdrop, various Privacy-Enhancing Technologies (PETs) are gaining attention. Notable PETs include local differential privacy, where users add noise to their personal data on their own devices before sending it to service providers; homomorphic encryption, which enables computation on encrypted data; and Trusted Execution Environments (TEE), which create protected isolated environments for data processing using hardware-based security mechanisms.

Our laboratory conducts practical research on applying these PETs to various types of data including location information, voice, and images. Specifically, we are researching the protection of location data using local differential privacy, implementing query processing mechanisms in TEE's isolated environments to achieve data structure-optimized query processing, and developing methods for different organizations to share and aggregate homomorphically encrypted data for analysis. Beyond these themes, after enrollment, students can conduct research aligned with their interests in specific data types and privacy-enhancing technologies.

References

Taisho Sasada, Yuzo Taenaka, and Youki Kadobayashi, “D$^2$-PSD: Dynamic Differentially-Private Spatial Decomposition in Collaboration with Edge Server”, IEEE Access, Oct 2024. (Accepted)

Taisho Sasada, Yuzo Taenaka, Youki Kadobayashi, “Oblivious Statistic Collection With Local Differential Privacy in Mutual Distrust”, IEEE Access, March 2023. DOI: 10.1109/ACCESS.2023.3251560

Tomoya Suzuki, Taisho Sasada, Yuzo Taenaka, Youki Kadobayashi, “MosaicDB: An Efficient Trusted / Untrusted Memory Management for Location Data in Database”, In The Sixteenth International Conference on Advances in Databases, Knowledge, and Data Applications (DBKDA 2024), February 2024.

Taisho Sasada, Nesrine Kaaniche, Maryline Laurent, Yuzo Taenaka, Youki Kadobayashi, “Differentially-Private Data Aggregation over Encrypted Location Data for Range Counting Query”, In 38th International Conference on Information Networking (ICOIN 2024), January 2024.