Secure Keyword Auction: Preserving Privacy of Bidding Prices and CTRs

We develop a secure keyword auction mechanism in which winners and payments are calculated without making bidding prices and CTRs public. First, we show that even if we utilize cryptographic techniques, the auctioneer can learn most bidding prices in existing keyword auction mechanisms, such as the Generalized Second Price (GSP) and the Vickrey-Clarke-Groves (VCG) mechanism. Thus we propose a new auction mechanism by introducing a simple but sufficient payment rule to prevent the leakage of bidding prices. The existing GSP is not strategy-proof, and neither is our new mechanism. However, possible manipulations are limited. We also propose a secure keyword auction scheme that securely realizes our mechanism by utilizing cryptographic techniques.

[1]  Katsumi Takahashi,et al.  Privacy Preserving Computations without Public Key Cryptographic Operation , 2008, IWSEC.

[2]  A. Yao,et al.  Fair exchange with a semi-trusted third party (extended abstract) , 1997, CCS '97.

[3]  Makoto Yokoo,et al.  Keyword Auction Protocol for Dynamically Adjusting the Number of Advertisements , 2008, 2008 IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology.

[4]  Benny Pinkas,et al.  Fairplay - Secure Two-Party Computation System , 2004, USENIX Security Symposium.

[5]  Michel Gendreau,et al.  Combinatorial auctions , 2007, Ann. Oper. Res..

[6]  Masayuki Abe,et al.  M+1-st Price Auction Using Homomorphic Encryption , 2002, Public Key Cryptography.

[7]  Tim Roughgarden,et al.  Algorithmic Game Theory , 2007 .

[8]  Moni Naor,et al.  Privacy preserving auctions and mechanism design , 1999, EC '99.