Arrow Research search

Author name cluster

Esther David

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

9 papers
2 author rows

Possible papers

9

IJCAI Conference 2017 Conference Paper

Enhancing Crowdworkers' Vigilance

  • Avshalom Elmalech
  • David Sarne
  • Esther David
  • Chen Hajaj

This paper presents methods for improving the attention span of workers in tasks that heavily rely on their attention to the occurrence of rare events. The underlying idea in our approach is to dynamically augment the task with some dummy (artificial) events at different times throughout the task, rewarding the worker upon identifying and reporting them. The proposed approach is an alternative to the traditional approach of exclusively relying on rewarding the worker for successfully identifying the event of interest itself. We propose three methods for timing the dummy events throughout the task. Two of these methods are static and determine the timing of the dummy events at random or uniformly throughout the task. The third method is dynamic and uses the identification (or misidentification) of dummy events as a signal for the worker's attention to the task, adjusting the rate of dummy events generation accordingly.

PDF Details

AIJ Journal 2014 Journal Article

On the choice of obtaining and disclosing the common value in auctions

  • David Sarne
  • Shani Alkoby
  • Esther David

This paper introduces a game-theoretic analysis of auction settings where bidders' private values depend on an uncertain common value, and the auctioneer has the option to purchase information that can eliminate that uncertainty. Therefore the auctioneer needs to decide whether to purchase the information, and if so, whether to disclose it to the bidders. Unlike prior work, the model assumes that bidders are aware of the auctioneer's option to purchase the external information but are not necessarily aware of her decision. The modeling of the problem as a Stackelberg game, where the auctioneer is the leader, is complicated by the fact that in cases where the auctioneer decides not to disclose the information, the situation is actually modeled as a version of Stackelberg game where the follower has potentially imperfect information about the leader's actions. Our analysis of the individual expected-benefit-maximizing strategies results in the characterization of the pure-strategy perfect Bayesian Nash equilibrium and proof of its existence for any setting. In addition, we introduce an algorithm for extracting the equilibrium as a function of the information cost, which is of great importance when the information is provided by a strategic information-provider. The analysis is also extended to deal with mixed-strategy perfect Bayesian Nash equilibrium and with noisy information. Overall, the analysis enables the demonstration of various model characteristics, including many non-intuitive properties related to the benefits of competition, the benefits in having the option of the auctioneer to purchase such information and the benefits encapsulated in the bidders' awareness of such an option.

Details DOI

AAMAS Conference 2007 Conference Paper

An Advanced Bidding Agent for Advertisement Selection on Public Displays

  • Alex Rogers
  • Esther David
  • TERRY R. PAYNE
  • Nicholas R. Jennings

In this paper we present an advanced bidding agent that participates in first-price sealed bid auctions to allocate advertising space on BluScreen – an experimental public advertisement system that detects users through the presence of their Bluetooth enabled devices. Our bidding agent is able to build probabilistic models of both the behaviour of users who view the adverts, and the auctions that it participates within. It then uses these models to maximise the exposure that its adverts receive. We evaluate the effectiveness of this bidding agent through simulation against a range of alternative selection mechanisms including a simple bidding strategy, random allocation, and a centralised optimal allocation with perfect foresight. Our bidding agent significantly outperforms both the simple bidding strategy and the random allocation, and in a mixed population of agents it is able to expose its adverts to 25% more users than the simple bidding strategy. Moreover, its performance is within 7. 5% of that of the centralised optimal allocation despite the highly uncertain environment in which it must operate.

PDF

ECAI Conference 2006 Conference Paper

Auction Mechanisms for Efficient Advertisement Selection on Public Displays

  • Terry R. Payne
  • Esther David
  • Nicholas R. Jennings
  • Matthew Sharifi

Public electronic displays can be used as an advertising medium when space is a scarce resource, and it is desirable to expose many adverts to as wide an audience as possible. Although the efficiency of such advertising systems can be improved if the display is aware of the identity and interests of the audience, this knowledge is difficult to acquire when users are not actively interacting with the display. To this end, we present BluScreen, an intelligent public display, which selects and displays adverts in response to users detected in the audience. Here, users are identified and their advert viewing history tracked, by detecting any Bluetooth-enabled devices they are carrying (e. g. phones, PDAs, etc.). Within BluScreen we have implemented an agent system that utilises an auction-based marketplace to efficiently select adverts for the display, and deployed this within an installation in our Department. We demonstrate, by means of an empirical evaluation, that the performance of this auction-based mechanism when used with our proposed bidding strategy, efficiently selects the best adverts in response to the audience presence. We bench-marked our advertising method with two other commonly applied selection methods for displaying adverts on public displays; specifically the Round-Robin and the Random approaches. The results show that our auction-based approach, that utilised the novel use of Bluetooth detection, outperforms these two methods by up to 64%.

Details
v2026.09.13