A Framework for Data Warehousing and Mining in Sensor Stream Application Domains - Evolving Application Domains of Data Warehousing and Mining - page 113

Chapter 6
A Framework for Data
Warehousing and
Mining in Sensor Stream
Application Domains

Nan Jiang
Cedarville University, USA

ABStrAct

The recent advances in sensor technologies have made these small, tiny devices much cheaper and
convenient to use in many different applications, for example, the weather and environmental monitor-
ing applications, the hospital and factory operation sites, sensor devices on the traffic road and moving
vehicles and so on. The data collected from sensors forms a sensor stream and is transferred to the server
to perform data warehousing and mining tasks for the end user to perform data analysis. Several data
preprocessing steps are necessary to enrich the data with domain information for the data warehous-
ing and mining tasks in the sensor stream applications. This chapter presents a general framework for
domain-driven mining of sensor stream applications. The proposed framework is able to enrich sensor
streams with additional domain information that meets the application requirements. Experimental stud-
ies of the proposed framework are performed on real data for two applications: a traffic management
and an environmental monitoring site.

IntroductIon

micro sensors, which can be integrated in a wireless
sensor network. These devices - Wireless Integrated
Network Sensors (WINS) - will enable fundamental
changes in applications spanning the home, office,
clinic, factory, vehicle, metropolitan area, and the
global environment.
Concerning the needs of the user for knowledge
discovery from sensor streams in these application
domains, new data warehousing, data mining tech-
niques have to be developed to extract meaningful,

Networks of thousands of sensors present a feasible
and economic solution to some of our most chal-
lenging problems, such as real-time traffic modeling,
weather and environmental monitoring, and military
sensing and tracking. Recent advances in sensor
technology have made possible the development
of relatively low cost and low-energy-consumption