Corelab Seminar

Aggelos Kiayias (U. Athens)
Fingerprinting codes

Abstract. A fingerprinting code is a probability distribution over sets of strings that accommodates an efficient tracing algorithm. This algorithm, given any string constructed adversarially out of a subset of codewords according to a constraint called the "marking assumption" is capable of recovering one of the "guilty" codewords with high probability. Fingerprinting codes have many applications in watermarking content and constructing encryption schemes for content distribution. In this talk, we discuss the problem of constructing fingerprinting codes and present some explicit constructions as well as explain how such fingerprinting codes can be used to construct multiuser encryption schemes.