Bs-Kgs: Blockchain Sharding Empowered Knowledge Graph Storage

Utilizing blockchain-based storage in knowledge graph construction is considered an alternative to ensure both security and traceability of files. Sharding is a promising solution to improve blockchain throughput by concurrently processing transactions. However, there are two key challenges in sharding, namely, difficulties in generating transaction allocation strategy and security threats. This paper addresses above issues and proposes a blockchain sharding-empowered knowledge graph storage (BS-KGS) scheme. To achieve optimal transaction sharding strategies, we construct a dynamic programming algorithm that considers multiple resource constraints. Moreover, our approach guarantees transaction security via implementing several methods, including Verifiable Random Function (VRF), asset mortgage, and Practical Byzantine Fault Tolerance (PBFT) consensus. Evaluation results have demonstrated that our method is efficient in transaction sharding.