Blockchain technology provides a decentralized, tamper-proof solution for cross-institutional education evaluation data sharing. This study proposes a cross-institutional education data sharing model that integrates blockchain, RSA cryptographic accumulators, and IPFS to achieve secure off-chain storage and efficient on-chain verification of education data. Using the RSA accumulator, multiple fingerprints of educational record data are aggregated into a single cryptographic accumulator for on-chain storage, enabling users to quickly verify the authenticity of individual data points via verifiable credentials. Attribute-based encryption (CP-ABE) is employed to protect the original data stored in IPFS, with students able to define their own access policies to ensure granular permission control. The experiment validated performance using the real-world dataset MOOCCube. RSA accumulator key slicing processed 200 key pairs in just 62.69 seconds, improving efficiency by 14.5% compared to the Slicing method. Hybrid encryption of 200 courses took 99.43 seconds, and smart contract management of 50 contracts took only 247.16 seconds, both significantly outperforming comparison schemes. Combined with Bloom filters to enable multi-keyword search, a 5-keyword search takes only 4.26 seconds, which is 67.1%–114.6% faster than the baseline scheme. A group-optimized consensus mechanism is designed to improve throughput, reaching a peak of 1023.61 TPS, which is 3.6 times higher than the ordinary scheme. Block recovery success rate reaches 100% when the replication factor c ≥ 2, and the direct recovery rate remains at 89.53% even when the node scale is expanded to 40 nodes. This model effectively improves the efficiency and scalability of cross-institutional education data sharing while ensuring data privacy and integrity.
