Achieving Strong Privacy in Online Survey.

Abstract

Thanks to the proliferation of Internet access and modern digital and mobile devices, online survey has been flourishing into data collection of marketing, social, financial and medical studies. However, traditional data collection methods in online survey suffer from serious privacy issues. Existing privacy protection techniques are not adequate for online survey for lack of strong privacy. In this paper, we propose a practical strong privacy online survey scheme SPS based on a novel data collection technique called dual matrix masking (DM²), which guarantees the correctness of the tallying results with low computation overhead, and achieves universal verifiability, robustness and strong privacy. We also propose a more robust scheme RSPS, which incorporates multiple distributed survey managers. The RSPS scheme preserves the nice properties of SPS, and further achieves robust strong privacy against joint collusion attack. Through extensive analyses, we demonstrate our proposed schemes can be efficiently applied to online survey with accuracy and strong privacy.