Research On Digital Copyright Detection Based On Blockchain

As for the traditional digital copyright detection,there is a trust problem between the two parties,and the security of copyright cannot be guaranteed.The main goal of this thesis is to combine blockchain technology to achieve trusted and safe digital copyright detection and better protect digital copyright.This thesis proposes a digital copyright detection architecture based on blockchain.The architecture uses the smart contract technology of blockchain 2.0 to detect copyright attribution,so as to solve the trust issue of the detection agency.The detection results are stored in the blockchain,to form a tamper-proof record.In addition,this thesis improves the local sensitive hash algorithm and improves the accuracy of similarity detection.This thesis mainly focuses on copyright detection technology based on blockchain.The main work includes two aspects:(1)This thesis proposes a protection architecture for digital rights detection,data transmission,and storage based on blockchain.Firstly,in order to ensure the trustworthiness of the detect results,the digital media providers(copyright owners)can verify the correctness of the results.If there is any objection,the smart contract performs arbitrate.Secondly,in order to ensure the security of copyright,the original digital media content is encrypted and stored in IPFS.The address where the encrypted content is stored and the detection results are recorded in the blockchain,thus forming a tamper-proof,traceable proof of existence.Based on the security analysis and experimental simulation,the security and reliability of the architecture are demonstrated.(2)To meet the needs of digital rights detection architecture,this thesis proposes an improved Simhash algorithm based on local sensitive hash algorithm.The improvement lies in the calculation method of feature weight and the generation method of fingerprint.The experiment is divided into two parts.The first part fits a linear regression model based on the vector space model algorithm and the improved Simhash algorithm and find out the threshold for the Simhash algorithm to determine the text similarity.The second part compares the accuracy of the improved Simhash algorithm,the Simhash algorithm,and the vector space model algorithm in the text-similarity calculation from three aspects: manual annotation,text length,and text type.The effects of text length and text type on accuracy were also analyzed.Experimental results show that the improved Simhash algorithm has high accuracy and is suitable for digital copyright detection architecture based on blockchain.