Recently, with the advancement of cloud computing technology, database owners can outsource their databases to the cloud to provide professional management of data at low cost. However, outsourcing the original database to the cloud server exposes all...
Recently, with the advancement of cloud computing technology, database owners can outsource their databases to the cloud to provide professional management of data at low cost. However, outsourcing the original database to the cloud server exposes all the private information in the database to the cloud server. The cloud server can exploit information obtained from outsourced databases and user's queries.
In this paper, we propose an efficient association rule mining method that protects the original database, user's queries, query results and guarantees accurate query results in a database outsourcing environment. To provide information protection, the original database is encrypted with ElGamal cryptosystem, and the user's query and support for each item is encrypted with Paillier cryptosystem. To provide efficient query processing on ciphertexts, we propose two comparison protocols that compares ciphertexts without exposing information. The SPET protocol can determine whether two ciphertexts are equal or not with only the operation cost of decrypting the ElGamal ciphertext once. The SC protocol can perform comparisons of two ciphertexts with only 1/8 of the cost compared to the ciphertext comparison protocol used in the previous studies. As a result of comparing the proposed P-FPG algorithm with DP-ARM which is the algorithm of the previous study, the performance is improved by 68%, 123% and 81% in the retail, Chess and BMS-1 datasets respectively.