Design and Implementation of an Intelligent Classification and Copyright Protection System for Academic Papers in Economic and Financial Fields Based on Blockchain

Abstract

Aiming at the pain points of difficult confirmation of academic copyright, insufficient refined management, and low efficiency of manual classification in the economic and financial domain, this paper develops an academic paper management system that integrates blockchain technology and machine learning-enabled intelligent classification. Supported by Hyperledger Fabric, IPFS distributed storage and hash encryption, the system leverages blockchain to accomplish copyright depository, differentiated permission management and full-link traceability. The XGBoost model is adopted to conduct three-level intelligent classification of papers (Grades A, B, and C), which supplies automated rule input for the permission system and automates the whole workflow encompassing classification, depository and permission configuration. The layered collaborative architecture of off-chain rule computation and on-chain data depository resolves the performance bottlenecks and iteration dilemmas triggered by uploading business rules onto the chain in conventional solutions. The system caters to two user roles: end-users and auditors. Experimental results indicate that the system achieves a classification accuracy of 82.1%, a depository latency of less than 200 ms, and a 100% success rate of permission verification, thereby realizing the trustworthy and intelligent management of academic copyright.