Enhance data usefulness in privacy protection under considering IoT measurement error

Recently, the use of big data has become crucial for the development of machine learning models and statistical analysis. Even if personal information is not directly revealed and only statistical results and models are published, there is still a risk that personal information can be inferred from them. Therefore, it's important to protect user privacy by excluding personally identifiable information and using a privacy protection method based on differential privacy, the standard metric used by major technology companies such as Google, Apple, and Microsoft. This method prevents personal identification by introducing noise into the actual values.

Typically, a higher level of privacy corresponds to more noise, which can reduce the usefulness of the data. There is a trade-off between the usefulness of the data and the level of privacy protection. Balancing data utility and privacy requires minimizing the added noise as much as possible.

Our research focuses on the inherent errors in measurements from IoT devices, with the goal of improving data utility while maintaining the level of privacy protection. Before adding noise for privacy protection, measurements from IoT devices already contain inherent errors. This suggests that adding noise for privacy protection could potentially introduce superfluous noise, affecting the desired level of privacy protection.

We propose a privacy protection method for both one-dimensional numerical data and two-dimensional location data. For numerical data, we introduce a method to determine the optimal approach by verifying whether it satisfies differential privacy standards through simulation. For location data, by introducing a new privacy standard, T-Geo-I, we present an innovative method that considers the measurement errors.

Authors;

Riho Isawa (Main)

-- The University of Electro-Communications, Master student

Yuichi Sei

-- The University of Electro-Communications, Professor

Yasuyuki Tahara

-- The University of Electro-Communications, Associate Professor

Akihiko Ohsuga

-- The University of Electro-Communications, Professor