Chapter 11: The Entity Validation System (EVS): A Conceptual Architecture


11.1 The Grid

Identity theft is the nation's fastest-growing crime, and it is not just terrorists who create identities to mask their presence; this crime is also committed by perpetrators who steal credit cards through various schemes on the Web or from trash dispensers. After 9/11, advances have been made to unify state driver's licenses and to regulate Social Security numbers more tightly. While these efforts will lessen the masking of perpetrators, data mining can also be used to validate the identify of individuals. In this chapter, a theoretical architecture for accessing multiple and disparate data sets will be described, the purpose of which is to validate a person's identity via machine-learning algorithms coupled with the experience of seasoned investigators.

The entity validation system (EVS) will develop predictive models that will support a wide range of field personnel and devices in confirming the identity of individuals. The concept is to provide data mining as a Web service for security deterrence purposes. The evolution of the Internet, wireless networks, and a number of information system standards provides the infrastructure for such a system. In the near future, this kind of service will be feasible using a pervasive network, linking systems and agents via a grid, providing real-time intelligence derived from autonomous data mining robots that are self-adjustable, evolutionary, and continuously learning, under the tutelage of experienced analysts.

The envisioned EVS would utilize remote data access via networks and an evolutionary real-time data mining system. The concept advanced here is one of an intelligence-gathering engine with many possible behavioral profiling applications and delivery options. For example, because of the technologies used to assemble this EVS, alerts could be delivered not only to humans but also to machines. At its core, the EVS would use the data mining technologies covered earlier in the book, such as agents and machine learning.




Investigative Data Mining for Security and Criminal Detection
Investigative Data Mining for Security and Criminal Detection
ISBN: 0750676132
EAN: 2147483647
Year: 2005
Pages: 232
Authors: Jesus Mena

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net