Privacy-Preserving Matching Protocols for Attributes and Strings

In this technical report we present two new privacy-preserving matching protocols for singular attributes and strings, respectively. The first one is used for matching of common attributes without revealing unmatched ones to each other. The second protocol is used to discover the longest common sub-string of two input strings in a privacy-preserving manner. Compared with previous work, our solutions are efficient and suitable to implement for many different applications, e.g., discovery of common worm signatures, computation of similarity of IP payloads. Key word: privacy-preserving attribute matching, longest common sub-string, elliptic curve cryptosystem