Database Reference
In-Depth Information
Analysis Summary
the current location, the time interval, and the air
temperature of detected environment during this
interval. All sensor nodes report to a single server.
The sensors are deployed on different places of the
botanical garden, collect and store the air tempera-
ture detected for a given time interval. The actual
air temperatures are taken as sensor readings that
are used as input for our simulation experiment.
In this section, we discussed how to setup and
implement a traffic management sensor stream
application in the domain-driven framework for
data warehousing and mining from sensor streams.
We described the process of the framework setup in
each implementing stage: the raw data stage, data
preprocessing stage, data warehousing and data
mining stage, and the query processing stage.
The performance of our proposed framework is
studied by means of simulation. Several different
data warehousing and data mining experiments
are conducted in order to evaluate the proposed
framework while using the Average Window
Size (AWS) approach, the linear interpolation
approach, the linear trend approach, the WARM
approach (Halatchev, 2005), and the CARM ap-
proach (Jiang, 2007).All these methods are applied
to our proposed framework to answer the user's
request for missing sensor value estimation. We
compared the estimation accuracy, running time
and memory space usage when applying each
method to our proposed framework.
Our performance study shows that the pro-
posed domain-driven framework can be applied
to different data warehousing and mining tasks.
For the traffic management sensor stream applica-
tion, the closed pattern based association mining
to estimate missing sensor data online is an area
worth to explore.
Framework Setup and Implementation
As discussed in the domain-driven framework for
data warehousing and mining from sensor streams,
raw data was collected from the environmental
monitoring sensor stream application and being
transformed into data that was ready to be pro-
cessed. The data was then enriched with the domain
information from the environmental monitoring
sensor network, and ready to be processed by
different data warehousing and mining processes.
The mining results were used to answer different
user queries at requests. Below, we describe the
process in each stage of the framework setup and
implementation for the environmental monitoring
sensor stream application.
Raw Data Stage
At the raw data stage, which is on the bottom
level of our proposed framework for sensor
stream application domain, raw data is collected
from the environmental monitoring sensor stream
application network. This information includes
each sensor's current location, sensor identifier,
the time interval for the reported value, and the
air temperature detected during this time interval.
The collected data was then transferred to the data
preprocessor to be cleaned, integrated and added
with domain information for the environmental
monitoring sensor stream application.
the environmental Monitoring
Sensor Stream Application
A second experiment was performed over sensor
data collected in the Huntington Botanical Garden
in Sam Marino, California. The sensor reports the
air temperature of several places in the gardens
for different time intervals.
The simulation data of the environmental moni-
toring application was collected in year 2008 at
various locations throughout the sensor network in
Huntington Botanical Garden. The data represents
Search WWH ::




Custom Search