Privacy-preserving efficient top-k spatial keyword search in outsourced cloud
University of New Brunswick
Convergence of technologies like cloud computing, mobile/wireless networks and smart phone technologies have led to the rapid development of Location-Based Services (LBS). There is a recent trend to migrate LBS to leverage the cloud for flexibility and cost savings; this poses a serious threat to the user privacy. In this thesis, we study the privacy-preserving top-k spatio-textual keyword search in an outsourced cloud. Most of the existing techniques make use of traditional symmetric or asymmetric encryption approaches. These schemes involve an additional overhead of decryption, not only affecting search performance adversely, but rendering them ineffective to be used in an honest-but-curious cloud server. To address the limitations of existing approaches, we propose a privacy-preserving scheme for top-k spatio-textual keyword search within a given distance in an outsourced cloud environment. To achieve scalability and security, we employ an improved homomorphic encryption technique over a composite order group. This technique enables a registered user to get top-k results without divulging the location information. We propose the use of hashbuckets to find the rough range of the spatial objects. Further, to reduce the search latency, we developed a secure index based on the I³ spatio-textual indexing approach. Extensive experiments and detailed analysis confirm the scalability, efficiency and security properties of the proposed scheme.