Jin Network Applies for Blockchain-Based Privacy Computing Patent, Achieving a Breakthrough in Data Security

Jin Network Applies for Blockchain-Based Privacy Computing Patent, Achieving a Breakthrough in Data SecurityFinancial News, December 2, 2024 Jin Network (Beijing) Digital Technology Co., Ltd

Jin Network Applies for Blockchain-Based Privacy Computing Patent, Achieving a Breakthrough in Data Security

Financial News, December 2, 2024 Jin Network (Beijing) Digital Technology Co., Ltd. has applied for a patent titled "Method, Device, Equipment, Medium, and Product for Privacy Computing Based on Blockchain," with publication number CN119046979A and application date August 2024, according to the State Intellectual Property Office. This patent technology addresses the challenges of data security and privacy protection during machine learning model training, representing a significant breakthrough in the field of privacy computing.

The patent abstract details the specific methods and processes. The core idea is to shift the machine learning model training process to the data source side, avoiding the direct transmission of raw training data, thereby effectively protecting data security and privacy. Specifically, the method includes the following key steps:

Step 1: Task On-Chain and Data Source Node Selection. First, the computing task used to train the machine learning model is uploaded to the blockchain network. This task includes the parameters and objectives required for model training, as well as requirements for participating nodes. The blockchain network selects suitable nodes as data source nodes based on task requirements. These nodes possess multiple copies of the training sample data needed for model training. This step ensures task transparency and traceability, and the distributed nature of the blockchain enhances task processing reliability and security.

Step 2: Local Model Training and Result On-Chain. After receiving the computing task, the data source node performs local model training. Only when a node determines that it possesses a sufficient number of training samples to meet the model training requirements will it initiate the training process. Crucially, the entire training process is completed in the local environment of the data source node, preventing the transmission and leakage of sensitive data. After training, the generated model training results are uploaded to the blockchain network. This ensures the authenticity and integrity of the training results while preventing potential tampering.

Step 3: Model Parameter Adjustment and Global Model Construction. The blockchain network collects model training results uploaded by multiple data source nodes. These results are treated as constraints in a system of equations. By solving this system of equations, the optimal adjustment values for the model parameters are obtained. These adjustment values represent the optimization direction for the global model parameters. This step leverages the computing power of the distributed network, integrating the training results from various nodes to ultimately obtain a more accurate and robust model.

Step 4: Global Model Verification and Deployment. Based on the obtained model parameter adjustment values, the system builds a new, global machine learning model. This global model then undergoes rigorous verification to ensure it meets preset performance indicators and training objectives. Only after passing verification is the trained model formally deployed and applied. This ensures the quality and reliability of the final model.

Jin Network's blockchain-based privacy computing method cleverly combines blockchain technology and privacy computing technology, enabling efficient machine learning model training while protecting data security and privacy. Traditional machine learning model training methods often require concentrating large amounts of sensitive data on a central server for processing, significantly increasing the risk of data leakage. Jin Network's patented technology solves this problem. By distributing the training process to multiple data source nodes and using blockchain technology to ensure data integrity and traceability, the method effectively reduces the risk of data leakage and improves data security and privacy protection.

The innovation of this patent technology lies in its unique model training process and data processing mechanism. It avoids the transmission of raw training data, instead uploading computing tasks and training results to the blockchain. Through distributed computing and collaborative optimization, it ultimately obtains a global, high-performance model. This not only improves data security and privacy protection but also enhances model training efficiency.

This technology has broad application prospects. In finance, healthcare, and the internet, large amounts of sensitive data need to be used for model training, and Jin Network's patented technology can provide a secure and reliable solution for these fields, promoting the application of artificial intelligence in various areas. For example, in finance, this technology can be used to train anti-fraud models, and in healthcare, it can be used to train disease prediction models without worrying about the leakage of patient privacy data.

This patent application demonstrates Jin Network's technical strength and innovation in the field of privacy computing. As data security and privacy protection become increasingly important, Jin Network's patented technology is expected to play a significant role in the future and contribute to the industrialization of privacy computing technology. It provides new ideas and methods for solving the problem of data security and privacy protection in machine learning model training, and lays the foundation for building a more secure and reliable data ecosystem. This innovation is not only significant for Jin Network's own development but also positively impacts the entire privacy computing industry. Through continuous technological innovation and exploration, Jin Network is expected to become a leader in the privacy computing field. We look forward to seeing this technology applied in more areas and making greater contributions to protecting data security and privacy. The successful application of this patent is undoubtedly great news for Jin Network and the entire Chinese technology community. It signifies that China's technological level in privacy computing has reached a world-leading position. In the future, as the technology continues to mature and improve, Jin Network's technology will surely bring security and convenience to more industries.