frequently it is hard to collect high-quality geosocial data.
Thus, it would be useful for organizations and for service
providers to be able to buy high-quality location data from
people. This requires having a marketplace in which people
could oer the data, with specified limitations, and buyers could buy data, according to their needs (Fig. 1). The
incentive for sharing location data can be money, as in Amazon Mechanical Turk (www.mturk.com), where people receive
money for executing small tasks. The incentive may also be
a variant of reward points that grant access to a service, or
provide some other benefit. Data may also be exchanged for
other data. However, when collecting data from a crowd,
exchange is problematic because most people have no need
for data of other people and thus there is a need for a marketplace where people could sell and buy data.
Recently there has been a growing interest in personal-
data marketplaces, and a few companies have been established to satisfy this need. However, they do not focus on the
specific demands of a marketplace for spatio-temporal data.
For example, a user, say Alice, may only be interested in recent location data of people traveling on the roads between
her home and her office, to estimate the traffic condition on
these roads. Another example is of an urban planning task
where the organization only needs information about the location of people who live in a certain neighborhood. The
challenge is to allow sellers and buyers to specify their needs
effectively and put the appropriate price tag on the data. In
this paper we elaborate on these questions.