Research and Application Framework for Trusted Circulation of Food Industry Data Based on Blockchain and Federated Learning

Food is a necessity for human survival. The circulation of data in the food industry possesses characteristics such as a long lifecycle, complex processes, heterogeneous and sensitive information sources, and susceptibility to leakage. The trend of security risks has shifted from frequent occurrences to emerging, sudden, and sporadic outbreaks. Blockchain, as a novel decentralized architecture and distributed computing paradigm, is gradually being applied in the field of food safety. Federated learning can achieve 'usable but invisible' data, improving data utilization and processing efficiency while allowing participants to benefit from data sharing. Blockchain ensures the security of the federated learning model sharing process by holding malicious model contributors and nodes accountable, preventing model data tampering. This paper first comprehensively analyzes the literature on blockchain and federated learning. Based on this analysis, combined with the current application status of blockchain in the food industry and the application ideas of federated learning in other industrial sectors, it proposes a research and application framework for trusted data circulation in the food industry based on blockchain and federated learning. The future development trends are also discussed. Research on trusted data circulation in the food industry can help improve the resilience of the entire food industry chain and the realization of the value of data elements.