Toward secure and effective data utilization in public cloud

Abstract

Cloud computing enables a promising paradigm of data service outsourcing, where data owners can avoid committing large capital outlays by economically storing their data to public data centers for the convenient management of data storage and utilization. Despite the tremendous benefits, outsourcing data to the commercial public cloud also opens the door for unsolicited data access in the cloud and beyond. Thus, enabling a secure and effective cloud data utilization service is of paramount importance. Given the large number of data users and huge amount of outsourced cloud data, this problem is particularly challenging as it is extremely difficult to also meet the practical requirements of performance, system usability, and high-level user searching experiences. This article investigates these challenges and in particular defines the problem of fuzzy keyword search over encrypted cloud data, which aims at accommodating various typos and representation inconsistencies in different user searching input for acceptable system usability and overall user searching experience, while protecting keyword privacy. In order to further enrich the application spectrum, we also demonstrate how the notion of fuzzy search naturally supports similarity search, a fundamental and powerful tool that is widely used in information retrieval. We describe the challenges that are not yet met by existing techniques, and discuss the research directions and possible technical approaches for these new search functionalities to become a reality.