The mobile computing landscape is witnessing an unprecedented number of devices that can acquire geo-tagged (aka location-based) data, for example, mobile phones, wearable sensors, in-vehicle dashcams, and IoT sensors. These devices can collect large amounts of data such as images, videos, movement parameters, or environmental measurements along with the data collectors' location data. However, we are giving away our location data to large Web search companies and social media companies for free. In addition, some of our smartphone apps gather our location data to sell to other companies for targeted advertising.
This data may be useful to third-party entities interested in gathering information from a certain location. For example, journalists may want to gather images around an event of interest for their newspaper; law enforcement may seek images taken soon before or after a crime occurred; and city authorities may be interested in travel patterns during heavy traffic.
An emerging trend is therefore to create data marketplaces where owners advertise their geo-tagged data objects to potential buyers, dubbed geo-market-places.1 The following paper describes a technique for computing the monetary value of a person's location data for a potential geo-marketplace. The authors postulate that buyers pay people for their location data, perhaps through a geo-marketplace, and it shows how to compute the monetary value of such locations (represented as GPS points).
The paper applies established decision theory and "value of information" (VOI) techniques to determine the worth of a GPS point, illustrated with three scenarios. The first scenario is for a buyer that wants to deliver advertisements to people who live in a certain geographic region. The buyer uses GPS points to gradually narrow down the location of each person's home, ideally gaining confidence in the location with each new point. The buyer must decide, for each user, which points to buy (if any) and make the ad delivery decision (that. is, decide to deliver the ad to that user or not). The dilemma for the buyer is that it cannot see the coordinates of the available GPS points until it actually pays for them, so it must make its purchasing decisions based only on other available data about the points, such as their timestamps. The paper's analysis gives a principled approach for the buyer to compute the expected value of unseen GPS data, leading to purchasing decisions that are demonstrably better than more simplistic approaches. By estimating the VOI of each point, the buyer can place a numerical value on each point giving its worth. The paper gives a second example scenario about using GPS data to estimate traffic speeds on roads, and a third scenario about predicting a person's future location. In all three scenarios, buying points with high VOI leads to more efficient use of the data resources by identifying the most valuable points for the task.
An emerging trend is to create data marketplaces where owners advertise their geo-tagged data objects to buyers, dubbed geo-marketplaces.
The authors introduce a general framework that can be adopted by any buyer or geo-marketplace requiring GPS data from users. The general framework allows the buyer to estimate the VOI of the offered GPS points, and it aids in making its purchasing and business decisions. The introduction of these principles means both buyers and sellers can set fair prices for GPS data, which today is still largely given away for no cost. Establishing these prices is a first step toward building a geo-marketplace, where sellers can be fairly compensated, and buyers can determine which data is most valuable for their application.
Geo-marketplaces, however, raise unique concerns. Publishing geo-tags reveal owners' whereabouts, which may lead to serious privacy breaches such as leakage of one's health status or political orientation. In addition, one must also protect the interests of buyers, and ensure they receive data objects satisfying their spatial requirements. Owners must be held accountable for their advertised data and not be able to change the geo-tag of an object after its initial advertisement. This can prevent situations where owners change geo-tags to reflect ongoing trends in buyers' interest. For example, when a certain high-profile event occurs at a location, dishonest owners may attempt to change their geo-tags closer to that location in order to sell their images at higher prices. Furthermore, the system must provide strong disincentives to prevent spam behavior, where dishonest participants flood the system with fake advertisements. These are possible future work for geo-marketplaces, some of which we discussed in a recent paper.1
1. Nguyen, K., Ghinita, G., Naveed, M., and Shahabi, C. A privacy-preserving, accountable and spam-resilient geo-Marketplace. In Proceedings of the 27th ACM SIGSPATIAL Intern. Conf. in Geographic Information Systems (Chicago, IL, USA, Nov. 5-8, 2019).
To view the accompanying paper, visit doi.acm.org/10.1145/3410387
The Digital Library is published by the Association for Computing Machinery. Copyright © 2020 ACM, Inc.
No entries found