Fair Data Transactions Across Private Databases

In this era of big data, data transactions are becoming more frequent and bringing new benefits for both data buyers and providers. However, there are no data transaction schemes to satisfy the special security and business requirements of data transactions. In this paper, we introduce the requirements in practical data transaction processes and compare the differences of requirements and techniques between data transaction and traditional data sharing. Next, we propose and implement the first data transaction scheme based on a newly defined concept probabilistic commutative encryption, which makes the decryption order of a probabilistic encryption algorithm and a deterministic encryption algorithm commutative. Since the security concerns of both sides can be achieved in a malicious model, our scheme is fair to both data buyers and data providers from a business perspective. Furthermore, our scheme allows data transaction to be partly outsourced to a semi-honest cloud storage platform.