Individual Fairness in Advertising Auctions Through Inverse Proportionality

@inproceedings{Chawla2022IndividualFI,
  title={Individual Fairness in Advertising Auctions Through Inverse Proportionality},
  author={Shuchi Chawla and Meena Jagadeesan},
  booktitle={ITCS},
  year={2022}
}
Recent empirical work demonstrates that online advertisement can exhibit bias in the delivery of ads across users even when all advertisers bid in a non-discriminatory manner. We study the design ad auctions that, given fair bids, are guaranteed to produce fair outcomes. Following the works of Dwork and Ilvento [11] and Chawla et al. [7], our goal is to design a truthful auction that satisfies “individual fairness” in its outcomes: informally speaking, users that are similar to each other… 

Figures from this paper

Individually-Fair Auctions for Multi-Slot Sponsored Search
TLDR
This work designs fair sponsored search auctions that achieve a near-optimal tradeoff between fairness and quality and provides a computationally efficient algorithm for computing payments for these auctions as well as those in previous work, resolving another open direction from [4].
The price of unfairness in linear bandits with biased feedback
TLDR
A Fair Phased Elimination algorithm that addresses the problem of sequential decision making with biased linear bandit feedback by establishing an upper bound on its worst-case regret, and obtaining gap-dependent upper bounds on the regret and revealing the existence of non-trivial instances where the problem is no more difficult than its unbiased counterpart.
Optimal data-driven hiring with equity for underrepresented groups
TLDR
A hiring policy that depends on the protected attribute functionally, but not statistically, is presented, and it is proved that, among all possible fair policies, the authors' is optimal with respect to the firm’s objective.

References

SHOWING 1-10 OF 26 REFERENCES
Toward Controlling Discrimination in Online Ad Auctions
TLDR
This work proposes a constrained ad auction framework that maximizes the platform's revenue conditioned on ensuring that the audience seeing an advertiser's ad is distributed appropriately across sensitive types such as gender or race, and presents an optimal auction mechanism for a large class of fairness constraints.
Multi-category fairness in sponsored search auctions
TLDR
This work proposes inter-category and intra-category fairness desiderata that take inspiration from individual fairness and envy-freeness, and investigates the "platform utility" achievable by mechanisms satisfying these desidersata.
Welfare Guarantees for Proportional Allocations
TLDR
This paper improves the lower bound of 26.8% on the price of anarchy over coarse-correlated and Bayes-Nash equilibria in the full and incomplete information settings, respectively and extends it to settings with budget constraints.
On the Efficiency of the Proportional Allocation Mechanism for Divisible Resources
TLDR
The Price of Anarchy (PoA) of the induced game under complete and incomplete information is studied, and it is proved that the PoA is exactly 2 for pure equilibria in the polyhedral environment.
Mechanism Design via Differential Privacy
TLDR
It is shown that the recent notion of differential privacv, in addition to its own intrinsic virtue, can ensure that participants have limited effect on the outcome of the mechanism, and as a consequence have limited incentive to lie.
Preference-informed fairness
TLDR
A new notion of preference-informed individual fairness (PIIF) is introduced that is a relaxation of both individual fairness and envy-freeness and can permit outcomes that are more favorable to individuals than any IF solution, while providing considerably more flexibility to the decision-maker.
Bidding strategies with gender nondiscrimination constraints for online ad auctions
TLDR
This work designs bidding strategies that advertisers can use to avoid emergent discrimination without having to modify the auction mechanism, and mathematically analyze the strategies to determine the additional cost to the advertiser for avoiding discrimination.
Fairness through awareness
TLDR
A framework for fair classification comprising a (hypothetical) task-specific metric for determining the degree to which individuals are similar with respect to the classification task at hand and an algorithm for maximizing utility subject to the fairness constraint, that similar individuals are treated similarly is presented.
Fairness Under Composition
TLDR
This work identifies pitfalls of naive composition and gives general constructions for fair composition, demonstrating both that classifiers that are fair in isolation do not necessarily compose into fair systems and also that seemingly unfair components may be carefully combined to construct fair systems.
Advertising for Demographically Fair Outcomes
TLDR
Under the assumption that a platform can reasonably enforce proportionality in spend across demographics, efficient exact and approximation algorithms for this problem are presented and simulation results on the datasets are presented to show the efficacy of these algorithms in achieving demographic parity.
...
...