Blockchain-Based Privacy-Preserving Alternative Credit Data Sharing
Yangyang Bao, Jianfei Sun, Xiaochun Cheng, Weidong Qiu, Liming Nie · IEEE Transactions on Computational Social Systems · 2025
In comparison to the lending data submitted by banks to credit bureaus under the traditional credit scoring paradigm, alternative credit data (such as social media activities and e-commerce consumption records) has increasingly demonstrated its significance in enhancing the accuracy of credit scores and addressing the issue of credit-invisible individuals in recent years. However, credit scoring model based on alternative credit data typically necessitates large-scale data circulation and may involve sensitive information, thereby raising concerns related to data security, user privacy, and data rights. Traditional cryptographic methods often encounter limitations in functionality, efficiency, flexibility, and traceability when addressing these issues. This article initially proposes a novel credit data sharing framework based on an alternative data cloud platform. Subsequently, based on this framework, a blockchain-based privacy-preserving alternative credit data sharing scheme is constructed. This scheme achieves efficient, privacy-preserving, and wildcard-supported attribute-based encryption (ABE) scheme through inner product operations, and implements a “two-level” access control by designing a keyword search mechanism in conjunction with the aforementioned scheme. Furthermore, a hybrid encryption mechanism is introduced to further enhance efficiency and security under high-frequency access scenarios. Security analysis and rigorous formal security reductions have been conducted to demonstrate the security of the proposed scheme. Comparative experimental results also indicate that the proposed scheme exhibits significant advantages in practicality compared with related schemes.