Multiparty Computation: Identifying the Consumers’ Willingness to Share Sensitive Automotive Data on MPC-­enabled Data Marketplaces

Multiparty Computation: Identifying the Consumers’ Willingness to Share Sensitive Automotive Data on MPC-­enabled Data Marketplaces: A Discrete Choice Modelling Approach

van Aalst, Christian (TU Delft Technology, Policy and Management)

Agahari, W. (mentor)
de Reuver, G.A. (mentor)
van der Voort, H.G. (mentor)

Delft University of Technology

Engineering and Policy Analysis

2021-09-02

Over the last several years, the volume of data generated by the Internet of Things (IoT) has expanded at a rapid pace around the world. These often very detailed sensor data could be utilized by many different parties in order to improve services. This same tendency is seen in the automotive sector, where lots of data are gathered by newer and smarter cars. However, despite the fast data collection, car businesses seldom utilize this potential value. One way to make better use of automotive data is to share it with others via data marketplaces. Furthermore, data gathered by increasingly smarter cars is often very sensitive data. By aggregating and analyzing these car data, other parties can learn a lot about individual car users which could be experienced as unpleasant. Therefore, Multiparty Computation (MPC), a fairly new technology that facilitates encrypted and anonymized data sharing, tries to overcome this problem in order to increase worldwide data sharing. Academics and businesses realize that these data marketplaces have huge potential but most of the research done on MPC is very technical. As Multiparty Computation is a fairly unknown and not yet massively adopted application and also not specifically on data marketplaces, it is unclear to what extent car users are willing to provide their car data. The scientific problem herein is to find the consumers' preferences in the main measurable data sharing factors which influence consumers' data sharing behavior on MPC-enabled data marketplaces. To do so, we will employ a explorative approach by the Discrete Choice Modeling (DCM) method. Conclusions are drawn on how to improve these data marketplaces by revealing the relative importance of general data sharing factors. This paper aims to explore the consumers' preferences of MPC-enabled data marketplace factors to reveal the consumers' willingness to share automotive GPS data on data marketplaces in order to enhance data sharing and increase (road suggestion) products and thereby overall innovation. The findings revealed that the factors risk of data disclosure and benefit stay far more important to consumers in data sharing on MPC-enabled data marketplaces. Data control and trust (in terms of a herding effect) were least important to consumers, although consumers feel they are losing control over their data in this online era. This knowledge can then be utilized to create a data marketplace in which data providers are more likely to join.

data marketplaces
Multiparty computation
data sharing
automotive
data control
risk of data disclosure
trust
benefit

http://resolver.tudelft.nl/uuid:9d4303cb-be3d-4964-b89e-c9736d840fcc

Student theses

master thesis

© 2021 Christian van Aalst