Blockchain Based Data Content Ownership Traceability And Trusted Sharing

Data sharing is an important step to gain maximum knowledge from previous work,and cross-domain and cross-industry data sharing is a necessary prerequisite for indepth mining of the value of big data.However,with the development of various data application technologies such as machine learning and federated learning,the development status of data sharing has not been significantly improved.The reason is that the key to restricting the development of data sharing technology lies in the privacy protection of the data circulation process.Most of the current common data sharing transaction environments need to rely on trusted third parties as an intermediary to endorse data transactions.However,there are obvious drawbacks in the centralized trading environment,such as,excessive participation of third-party entities,single point of failure and other drawbacks.Although there is currently a decentralized data sharing transaction environment based on blockchain,and the key mechanism is used to realize the data access policy control and the privacy protection of transaction entity information,the privacy protection of content information in the data transaction process is ignored.Data sharing is not a one-time transaction,and it is necessary to ensure that the transaction between the data owner and the data user is equal.However,in the existing decentralized trading environment,after the data is traded,the ownership of the data is transferred,and the income of the data owner will not be guaranteed,so the peer-to-peer transaction will not be possible.Therefore,this thesis realizes the peerto-peer transaction of data,authenticity proof,and ownership traceability under the premise of ensuring the privacy of data content.The research work of this thesis consists of three parts:(1)A data labeling algorithm based on Merkle tree is proposed,which provides the authenticity proof of the content for the data sharing transaction process,converts the data content into a public data identifier,hides the real data content,and realizes the privacy protection of the data content.(2)A data ownership traceability algorithm based on random oracles is proposed,which provides ownership proof for the data content of the transaction,and realizes the ownership protection in the process of data transaction with the data content identification provided by the data labeling algorithm.(3)Designed and implemented a trusted data transaction environment based on blockchain privacy protection,implemented data labeling algorithm and data ownership traceability algorithm through smart contracts,and arbitrated to verify the authenticity and ownership of transaction data content.In this thesis,IPFS and Ethereum are used to achieve a decentralized data transaction environment,and smart contracts and Ethereum kecceak256 are used to simulate the implementation of data labeling algorithms and random oracles,and arbitration smart contracts are used to realize content verification and data ownership in the process of data transaction and traceability.The experimental results show that the data content proof and index provided by the Merkle tree-based data labeling algorithm have lower computational cost and running time efficiency compared with the zero-knowledge proof algorithm.In this thesis,the Ethereum keccak256 is used with random numbers to realize the encryption computing for the privacy information in the transaction process.Compared with the traditional Advanced Encryption Standard(AES)128 and 256,the execution overhead of smart contracts is reduced,and the computing efficiency is significantly improved.