A Technical Concept for Pay-per-Use in Geomarketing Services
this moment, GI providers sell geographical and statistical data in
packages, and Geomarketing functions are performed locally in complex
systems. Smaller companies can not afford such systems, and they have
information need only from time to time. A solution which offers
geomarketing information as a service over the net is the answer:
minimal initial investment, and payments comparable to the use of the
data. This thesis shall investigate and develop a technical concept for
a pay-per-use mechanism for an electronic geodata commerce.The focus is
on the problem of ownership in data: the user will receive data in an
encrypted form (and owned by the service provider); for each use of the
data, the user must pay. The provider requires a secure method to
assure that the user can not use the data without payment.
analysis the user performs with the data and using software from the
provider entails the functionality of SQL with some extra procedures to
allow spatial analysis. The data provided in build (eg. on a CD-ROM)
will be augmented by updates to them, but the user has also the ability
to combine the providers data with his own data for analysis.
can formally model and show how the stream of data flows in this
environment, and that any time the customer uses the data the provider
gets a message. In this model, we can proof that all streams are secure
and whenever data from the provider is used, payment data is collected,
and that the providers data cannot leave the environment.
with studying the GISMO project, where this topic is embedded in.
Describe informally the setting, the data flow, the required protocol
mechanisms, and find out the related objects and methods (or functions).
the informal description: develop a formal model of the stream of data
flows in this environment in HASKELL. The formal model shall prove the
hypothesis without going into implementational details.
third step implement the model with JAVA and test your implementation
with real data sets (from GISMO partners). Re-use as far as possible
libraries available. As a result, you should be able to demonstrate
your specified model (perhaps with limited functionality) in relevant