Database Reference
In-Depth Information
6. Conclusions and Summary
While RFID is a relatively old technology, its use for large scale ap-
plications has proliferated in recent years. This is because of technolog-
ical advances in manufacturing, which have made the tags smaller and
cheaper. The ability to manufacture a tag at less than 5 cents (per tag)
has allowed their widespread use in a cost effective way. RFID data
brings numerous challenges with it for the purposes of mining and anal-
ysis. RFID data is inherently noisy and redundant because of missed
tag readings, or multiple readings of the same tag from different readers.
Therefore, techniques need to be designed in order to make the process of
reading more robust and reduce the redundancy in the underlying data.
The massive volume of the RFID data also makes the process of ware-
housing and querying the RFID data much more challenging. Therefore
methods need to be designed in order to represent the RFID warehouse
in terms of the aggregated views of RFID items which typically move
together. These aggregated views greatly improve the eciency of data
storage and querying. RFID data can be useful in detecting important
semantic events from the underlying data streams. The existing work
in active databases and sensor stream event detection can be further
extended in a variety of ways to make it suitable to the RFID scenario.
For example, methods have recently been designed for event processing
in uncertain RFID data streams.
RFID data naturally leads to a number of privacy challenges, because
of the association of people with tags, and the likelihood of monitoring
people's location with such tags. The privacy issues with RFID data
arise
both
during data collection and management. A number of methods
such as the kill command, cryptographic protocols, and blocker tags have
been designed for privacy protection during data collection. In addition,
a number of methods for physical access control have been developed for
preserving personal privacy during data management.
References
[1] R. Adaikkalavan, S. Chakravarthy. Snoopib: interval-based event
specification and detection for active databases.
Data and Knowl-
edge Engineering
, 59(1): pp. 139-165, 2006.
[2] C. C. Aggarwal. Data Streams: Models and Algorithms,
Springer
,
2007.
[3] C. C. Aggarwal, P. S. Yu. Privacy-Preserving Data Mining,: Models
and Algorithms,
Springer
, 2008.
