Font Size: a A A

Research On Cross-domain Data Sharing Technology Based On Blockchain

Posted on:2022-04-09Degree:MasterType:Thesis
Country:ChinaCandidate:J M ChenFull Text:PDF
GTID:2518306524489914Subject:Master of Engineering
Abstract/Summary:PDF Full Text Request
The traditional data management method is that each management domain independently manages the data in its own jurisdiction,which is not conducive to information sharing in the era of big data.Cross-administrative data sharing has become an urgent need.Traditional data sharing requires a third-party organization to uniformly manage and maintain data.There is a problem with this model.The data provider who needs to participate in data sharing absolutely trusts that the third-party data management organization is reliable.Unfortunately,The data provider cannot fully trust the third-party organization,which makes the data provider lack the subjective initiative of data sharing,which hinders the further development of data sharing.In order to solve the above-mentioned problems of the traditional data sharing model,this thesis proposes to build a new type of decentralized cross-administrative data sharing system by using the decentralized and non-tamperable characteristics of the blockchain.The main tasks are as follows:(1)In order to achieve cross-administrative data sharing,this thesis aims at the problem of inconsistent data naming in the process of cross-administrative data sharing.Starting from improving the efficiency of data access,this thesis proposes two solutions.For small-scale data sharing,the Unified Naming Model for Small-scale Data(UNSSD)is proposed.In this model,the Small-scale Data Unified Name(SDUN)algorithm is proposed to solve the problem of small-scale data sharing.The problem of inconsistent data naming in the process of large-scale data sharing;for large-scale data,the Unified naming model for large-scale data(UNLSD)is proposed,which combines Word2 vec and knowledge graphs to solve the large-scale data sharing process Inconsistent data naming issues.Finally,experiments are carried out on the two models to verify the effect of the unified naming of the data.(2)Based on the block chain storage data,for small-scale data,a new data structure based on Merkle DAG and union search set is designed,namely,the Union Find-Merkle DAG(Union Find-Merkle DAG);for large-scale data Data,proposes a data storage method based on blockchain and distributed file system.Design three kinds of smart contracts and deploy them to the blockchain to realize the intelligent storage,access and identity verification of data in the blockchain.(3)Design a cross-administrative domain data sharing system that combines the models and data storage structures mentioned in(1)and(2)to provide users with efficient and safe cross-administrative data sharing.The experimental results show that the cross-administrative domain data sharing system and data unified naming model designed in this thesis can realize efficient and safe data sharing in different management domains without the management and scheduling of a third-party management center.
Keywords/Search Tags:unified data naming model, blockchain, data sharing, word2vec, knowledge graph
PDF Full Text Request
Related items