A COPYRIGHT DETERMINATION SCHEME BASED ON META LEARNING IN WEB3

Authors

  • Hye-young Kim Author

DOI:

https://doi.org/10.46121/pspc.54.2.56

Keywords:

Meta-Learning, Few-Shot Copyright Detection, Blockchain-Based Verification, Web3 Architecture, NFT Ownership Authentication

Abstract

These The rapid adoption of Web3 technologies has transformed digital content distribution through decentralized storage, blockchain infrastructures, and non-fungible tokens (NFTs). However, existing blockchain-based copyright systems primarily ensure ownership registration and immutability, while lacking reliable mechanisms for semantic infringement detection and adaptive threat resilience. Moreover, conventional deep learning–based copyright identification approaches require large-scale labeled datasets and exhibit limited generalization to emerging content types, posing challenges in dynamic and adversarial web environments.

In this paper, we propose a secure copyright determination scheme based on meta-learning for Web3 ecosystems. Our proposed scheme integrates a meta-trained embedding network capable of few-shot adaptation with a blockchain-anchored verification mechanism. By leveraging metric-based meta-learning, the system rapidly generalizes to unseen copyright categories under limited labeled data and performs similarity-based infringement detection in a robust embedding space. To ensure integrity and non-repudiation, the extracted feature representations are transformed into cryptographic hashes and recorded via smart contracts, enabling tamper-resistant, transparent, and auditable ownership validation within a decentralized architecture.

Comprehensive experimental evaluations validate the effectiveness, robustness, and operational efficiency of the proposed scheme in data-constrained and adversarial scenarios. The results demonstrate statistically significant improvements in detection accuracy and adaptation speed compared with conventional baselines, while maintaining low transaction latency and controlled gas consumption. Security analysis further confirms resistance against data tampering and unauthorized ownership manipulation. Our proposed scheme provides a scalable, secure, and trustworthy solution for copyright determination in decentralized web platforms, NFT marketplaces, and metaverse-based digital services.

Downloads

Published

2026-06-18