Enabling Cooperative Privacy-preserving Personalized search in cloud environments

Abstract

With the booming of the information society, individuals and companies have been generating huge amounts of data every second. As a result, as a new storage tool, cloud computing has provided great convenience for data storage and application. However, it is worth noting that security and privacy issues in cloud computing have hindered its further development. To solve this problem, researchers proposed to encrypt the sensitive data. However, the encrypted data brought some follow-up issues such as an increased computational overhead and information retrieval inconvenience. In this paper, to enable users to obtain the most satisfactory results in searching outsourced encrypted data and reduce the computational overhead of cloud servers as well, the Cooperative Privacy-preserving Personalized Search (CPPS) scheme in cloud environments is proposed, which makes use of matrix encryption to ensure the privacy of the user. Combining the search of multiple users has greatly reduced the computing cost of the cloud server and guaranteed the accuracy of the user’s personalized search results at the same time. Experiments on the Yelp dataset indicate that the CPPS scheme has low index construction overhead, low computational overhead and communication overhead while providing users with personalized search results.