Year: 2024
Type: conference-paper
Abstract: Using machine learning modeling for user identification is a hot research direction in the field of computer science and application. However, the collection and use of a large amount of personal data by machine learning may pose risks of privacy leakage. This paper utilizes federated learning technology for joint user identification modeling, where participating parties do not need to expose their respective user data, thus protecting the privacy of personal user data. At the same time, combining blockchain technology for the notarization and sharing of model parameters ensures the authenticity of the parameters and the traceability of the model training process, enhancing mutual trust among participating parties and improving the auditability and transparency of joint modeling.
Authors: Quan Yuan, Yunlong Xie, Hanning Zhang, Jun Song
Institution: China United Network Communications Group (China)
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W4410857896