Implementation of a Blockchain-enabled Federated Learning Model that Supports Security and Privacy Comparisons
Xinyu Guo · 2022 IEEE 5th International Conference on Information Systems and Computer Aided Education (ICISCAE) · 2022
The rapid development of the information technology era has resulted in increased awareness of data privacy by the general public, and the corresponding laws and regulations regarding data protection are gradually enacted, which significantly impacts the development of data-dependent artificial intelligence. The federated learning model is used to fully protect users' privacy and maximize the value of the data created through the process of data use. However, the traditional federated learning algorithm is easily intercepted during use, resulting in data leakage and distortion. Suppose the federated learning model can be combined with the blockchain. In that case, the blockchain structure can be fully utilized, and the characteristics of the blockchain provide credible security to the federated learning algorithm, which will ensure that the algorithm fully protects the user's data privacy. This paper presents a federated learning security and privacy model enabled by blockchain technology and analyzes its data algorithm architecture in detail. This paper aims to optimize the above algorithm to ensure that it can be used normally for the protection of user data.