A privacy-preserving multi-keyword search approach in cloud computing

Cloud computing provides the users with the ability to outsource their data to a third-party cloud storage for cost-effective management of resources and on-demand network access. However, outsourcing the data to a third-party location may raise concerns about data privacy. To maintain the user's privacy, users tend to encrypt their sensitive data before outsourcing it. Encrypting the data will preserve its privacy, but at the same time, it makes the searching process for a specific keyword a time-consuming and challenging process, mainly if the encryption key is not provided. On the other hand, the data owner should be able to perform multiple keyword searches to retrieve specific documents that are relevant to the search query. This paper proposes a new privacy-preserving multi-keyword search approach for the cloud outsourced data. The objective of the proposed approach is to allow the data owners and the authorized users to retrieve the most relevant data with minimum computation and communication overhead, and reduced false positives (irrelevant documents) and searching time. To evaluate the proposed approach, the NSF research dataset is used. Results demonstrate that the proposed method achieves better searching time and overall performance of the cloud environment regarding computation and communication overhead as well as false positives in comparison with other approaches.