Rank and Impression Estimation in a Stylized Model of Ad Auctions

In order to help inform bidding decisions, advertisers in sponsored search auctions often create estimates of expected clicks and costs using aggregate information, such as the average positions and costs that their ads received on previous days. While aggregate information is typically what search engines reveal, it has been shown that using such data can produce biased estimates [1]. In this work, we construct a disaggregated model of positions and impressions from aggregate data. Previous work has demonstrated the feasibility and benefits of disaggregating ad auction data [2]. We extend the previous approach by formulating the disaggregation process as two intertwined problems, rank and impression estimation. We solve these problems using dynamic, mathematical, and constraint programming techniques. We evaluate the merits of our solution techniques in a simulated ad auction environment, the Ad Auctions division of the annual Trading Agent Competition.

[1]  Johannes Bisschop,et al.  AIMMS - Optimization Modeling , 2006 .

[2]  Brendan Kitts,et al.  Optimal Bidding on Keyword Auctions , 2004, Electron. Mark..

[3]  Aranyak Mehta,et al.  AdWords and generalized on-line matching , 2005, 46th Annual IEEE Symposium on Foundations of Computer Science (FOCS'05).

[4]  Michael P. Wellman,et al.  Designing an Ad Auctions Game for the Trading Agent Competition , 2009, AMEC/TADA.

[5]  Zizhuo Wang,et al.  A unified framework for dynamic pari-mutuel information market design , 2009, EC '09.

[6]  David M. Pennock,et al.  Revenue analysis of a family of ranking rules for keyword auctions , 2007, EC '07.

[7]  Michael P. Wellman,et al.  Autonomous bidding agents - strategies and lessons from the trading agent competition , 2007 .

[8]  Ashish Goel,et al.  Truthful auctions for pricing search keywords , 2006, EC '06.

[9]  Elizabeth Sklar,et al.  Agent-Mediated Electronic Commerce and Trading Agent Design and Analysis, AAMAS 2007 Workshop, AMEC 2007, Honolulu, Hawaii, May 14, 2007, and AAAI 2007 Workshop, TADA 2007, Vancouver, Canada, July 23, 2007, Selected and Revised Papers , 2009, International Workshop on Agent-Mediated Electronic Commerce and Trading Agents Design and Analysis.

[10]  Peter Stone,et al.  Adapting Price Predictions in TAC SCM , 2007, AMEC/TADA.

[11]  Nicole Immorlica,et al.  Multi-unit auctions with budget-constrained bidders , 2005, EC '05.

[12]  Peter Stone,et al.  The 2007 TAC SCM Prediction Challenge , 2008, AMEC/TADA.

[13]  Victor Naroditskiy,et al.  A Knapsack-Based Approach to Bidding in Ad Auctions , 2010, ECAI.

[14]  Glenn Ellison,et al.  Position Auctions with Consumer Search , 2007 .

[15]  Victor Naroditskiy,et al.  RoxyBot-06: Stochastic Prediction and Optimization in TAC Travel , 2009, J. Artif. Intell. Res..

[16]  Peter Stone,et al.  TacTex09: a champion bidding agent for ad auctions , 2010, AAMAS.

[17]  Benjamin Leblanc Holds a Bs Optimal Bidding on Keyword Auctions , 2010 .