Data Sharing Of Multisource Heterogeneous Database

The objective of this thesis is to introduce a novel technique for the sharing of data from multiple,diverse databases.We establishes a knowledge graph for the data of multi-source heterogeneous database,and uses deep learning method to mine the relationship between data.To prevent redundancy of data,Named Data Network research has been used to assign names to the entities in the knowledge map.To lessen the reliance on storage capacity locally,IPFS distributed storage is employed to store these data.We also use blockchain technology to control the reliability of knowledge sources and to a certain extent decentralize them.Through experiments,we have realized the preliminary model of the research on the data sharing model of multi-source heterogeneous databases,and implemented some functions,including search and upgrade.The search function is the display result of this study,and the upgrade function is to meet and adapt to various technologies that are constantly updated,such as deep learning methods.Our experiment has led us to the conclusion that utilizing knowledge map,deep learning and blockchains can enable the sharing of multi-source heterogeneous data without the third-patry,while also guaranteeing data privacy.The main contribution of this thesis is to provide new ideas for databases with multi-source heterogeneous databases.The innovation lies in the use of blockchain and other technologies to achieve the reliability of knowledge graph sources.