Optimal Auctions Capturing Constraints in Sponsored Search

  • Authors:
  • Esteban Feuerstein;Pablo Ariel Heiber;Matías Lopez-Rosenfeld;Marcelo Mydlarz

  • Affiliations:
  • Departamento de Computación, FCEyN, Universidad de Buenos Aires, Argentina;Departamento de Computación, FCEyN, Universidad de Buenos Aires, Argentina;Departamento de Computación, FCEyN, Universidad de Buenos Aires, Argentina;Yahoo! Research, Santiago, Chile

  • Venue:
  • AAIM '09 Proceedings of the 5th International Conference on Algorithmic Aspects in Information and Management
  • Year:
  • 2009

Quantified Score

Hi-index 0.01

Visualization

Abstract

Most sponsored search auctions use the Generalized Second Price (GSP) rule. Given the GSP rule, they try to give an optimal allocation, an easy task when the only need is to allocate ads to slots. However, when other practical conditions must be fulfilled ---such as budget constraints, exploration of the performance of new ads, etc.--- optimal allocations are hard to obtain. We provide a method to optimally allocate ads to slots under the practical conditions mentioned above. Our auctions are stochastic, and can be applied in tandem with different pricing rules, among which we highlight two: an intuitive generalization of GSP and VCG payments.