Privacy-preserving polynomial interpolation and its applications on predictive analysis
© 2020 Elsevier Inc. Privacy-preserving polynomial interpolation refers to a process which requires two parties to jointly finding out a polynomial over their private coordinate pairs. Unfortunately, the existing general approach remains impractical. To date, no practical solution to privacy-preserving polynomial interpolation exists. In this paper, we aim to fill this gap by presenting an efficient solution to enable this process. To this end, we first transform the privacy-preserving polynomial interpolation into privacy-preserving calculation of function values, and design a succinct privacy-preserving scalar product protocol. Then, we tackle the original problem by employing Lagrange interpolation in combination with our privacy-preserving scalar product protocol. Finally, we offer some application examples of how our protocol can be used to conduct privacy-preserving predictive analysis.