Seminario Ashkan Eshghi

Ashkan Eshghi (Warwick Business School) will present his paper “Algorithms to the Rescue: Market Mechanisms for Consensual Trading of Unbiased Individual Data.”

Abstract

The online economy has relied on collecting and monetizing users’ individual data in exchange for tools and services. A lack of transparency and the absence of proper compensation mechanisms have gradually eroded data quality in this market, giving rise to a new generation of platform-mediated data markets that aim to explicitly reimburse data subjects in return for their individual data. Moreover, these platforms can create data markets for direct collection of data from users in contexts like surveys and healthcare research. Such platforms, however, use either centralized-optimization or ¿xed-compensation mechanisms, which lead to expensive and/or biased samples for data buyers. In this paper, we present an algorithmic market mechanism approach which employs an incentive-compatible compensation mechanism in conjunction with a novel sampling method to enable such platforms to provide data buyers with low-cost, unbiased samples, while properly compensating data subjects for their loss of privacy. We illustrate the superior performance of our market mechanism against current approaches. We ¿nd that our approach outperforms the ¿xed-compensation approach, and even if the platform has access to partial information about data subjects’ privacy concern, in practice, they are better o¿ foregoing this information and using our approach instead. Finally, we provide insights about the trade-o¿s between bias and cost in data samples, and about the implications of sample size and anonymity.