TY - GEN
T1 - Information elicitation for Bayesian auctions
AU - Chen, Jing
AU - Li, Bo
AU - Li, Yingkai
N1 - Funding Information:
The first author thanks Matt Weinberg for reading a draft of this paper and for helpful discussions. The authors thank Constantinos Daskalakis, János Flesch, Hu Fu, Pinyan Lu, Silvio Micali, Rafael Pass, Andrés Perea, Elias Tsakas, several anonymous reviewers, and the participants of seminars at Stony Brook University, Shanghai Jiaotong University, Shanghai University of Finance and Economics, Maastricht University, MIT, and IBM Thomas J. Watson Research Center for helpful comments. This work is partially supported by NSF CAREER Award No. 1553385. A full version of this extended abstract is available at https://arxiv.org/abs/1702.01416.
Publisher Copyright:
© 2018, Springer Nature Switzerland AG.
PY - 2018
Y1 - 2018
N2 - In this paper we design information elicitation mechanisms for Bayesian auctions. While in Bayesian mechanism design the distributions of the players’ private types are often assumed to be common knowledge, information elicitation considers the situation where the players know the distributions better than the decision maker. To weaken the information assumption in Bayesian auctions, we consider an information structure where the knowledge about the distributions is arbitrarily scattered among the players. In such an unstructured information setting, we design mechanisms for auctions with unit-demand or additive valuation functions that aggregate the players’ knowledge, generating revenue that are constant approximations to the optimal Bayesian mechanisms with a common prior. Our mechanisms are 2-step dominant-strategy truthful and the revenue increases gracefully with the amount of knowledge the players collectively have.
AB - In this paper we design information elicitation mechanisms for Bayesian auctions. While in Bayesian mechanism design the distributions of the players’ private types are often assumed to be common knowledge, information elicitation considers the situation where the players know the distributions better than the decision maker. To weaken the information assumption in Bayesian auctions, we consider an information structure where the knowledge about the distributions is arbitrarily scattered among the players. In such an unstructured information setting, we design mechanisms for auctions with unit-demand or additive valuation functions that aggregate the players’ knowledge, generating revenue that are constant approximations to the optimal Bayesian mechanisms with a common prior. Our mechanisms are 2-step dominant-strategy truthful and the revenue increases gracefully with the amount of knowledge the players collectively have.
KW - Distributed knowledge
KW - Information elicitation
KW - Removing common prior
UR - http://www.scopus.com/inward/record.url?scp=85053281347&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-99660-8_5
DO - 10.1007/978-3-319-99660-8_5
M3 - Conference article published in proceeding or book
AN - SCOPUS:85053281347
SN - 9783319996592
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 43
EP - 55
BT - Algorithmic Game Theory - 11th International Symposium, SAGT 2018, Proceedings
A2 - Deng, Xiaotie
PB - Springer Verlag
T2 - 11th International Symposium on Algorithmic Game Theory, SAGT 2018
Y2 - 11 September 2018 through 13 September 2018
ER -