SGPM: a privacy protected approach of time-constrained graph pattern matching in cloud

Graph pattern matching (GPM) is an important operation on graph computation. Most existing work assumes that query graph or data graph is static, which is contrary to the fact that graphs in real life are intrinsically dynamic. Therefore, time-constrained graph pattern matching has been introduced. However, querying time-constrained graph pattern in a temporal graph is not an easy work because of the high computation complexity. Query outsourcing, with the help of cloud computing, is adopted in this paper. Outsourcing of data storage and query becomes increasingly popular due to the prevalence of cloud computing. However, sensitive data need to be encrypted before outsourcing for various privacy concerns. To execute queries over encrypted data is a very challenging problem and has received much attention recently. However, most existing approaches only support limited kinds of queries, moreover, they cannot be completely outsourced. In this paper, a Somewhat Homomorphic Encryption (SHE) approach is adopted to design a protocol which enables general queries on encrypted data and query outsourcing as well. The key issue in the paper is to (1) enable the query provider to filter data rows with homomorphic encrypted result of comparison operators; (2) completely relieve the data owner of partaking of the process of queries. An effective dualcloud protocol (DCP) which enables the cloud to understand the result of comparisons though homomorphic encrypted values is proposed. Moreover, The efficiency of the baseline approach is greatly improved by packing and GPU-accelerating technologies. Experimental study shows that the optimized approach outperforms the baseline approach and the recently reported similar approaches