| Terror Hunt: Warehousing and Mining Data |
|
The
official announcement of every terrorist attack invariably concludes with
a stock phrase, “Ghatna sthal ke nikat suraksha ke prabandh our kare
kar diye gaye hain (security arrangements around the place of incident
have been further strengthened).” However each incident is succeeded by
another attack, to be followed by a similar announcement. Terror strikes
at the place and time of its choosing. Yet there is a pattern and that
pattern essentially lies in the behaviour of the terrorist and modus
operandi of the parent organization. Future
attacks cannot be predicted, but past ones can be catalogued, stored and
retrieved in near real-time. “Data archiving” and “data mining”
are scientific tools and methologies, the former for gathering, sifting,
hoarding and storehousing data, and the latter for displaying duly
processed critical information to the decision-maker, whenever the need
arises. Its hallmarks are virtuality and event-retrieval potential, and
its purpose is to match, locate and track saboteurs, hijackers and
terrorists. In
April 1999, Applied Systems Intelligence Inc was selected by the US Air
Force to develop innovative information technology for a Global
Information Base to “store global awareness information,” besides
providing information services for dynamic planning and execution of
operations. The software developed by the firm is called KARNAC, short for
Knowledge-Aided Retrieval in Activity Context. It is highly versatile, and
is anchored in a group of technologies and decision support and database
management systems. It is designed to detect and identify impending
terrorist operations and similar missions. It
is well known that Al Qaida terrorists and others of their ilk hunt
for information on the Internet, often leaving valuable clues while
surfing and communicating. Therefore it is logical to look for and pursue
them in their haunt rather than go on hunting missions. Adam Pasik writes
in “Sifting through Data to Detect New Attacks,” (infowar.com), “The
problem is that intelligence and law-enforcement agencies are searching
the world’s biggest haystack – untold exabytes, or quintillions of
bytes of data stored on computers across the globe – to uncover a few
dangerous needles.” At
the time of the September 11 attacks, there was a plethora of helpful
scraps of information available e.g. e-mail intercepts, telephone calls,
car rentals, airline reservations, financial transactions, casino
winnings, Immigration records and much more. During the attack on our
Parliament, the terrorists left behind pertinent information such as a
laptop, which has reportedly been sent to Microsoft for analysis, and
vital information about the terrorists’ hawala (money laundering)
links and ISI connections garnered from cellular numbers called by the
terrorists. Whereas security, intelligence and law-enforcement agencies
work in tandem in the Critical
event detection, information retrieval and knowledge-based technologies,
products and systems are available off-the-shelf, and are widely used in
the commercial world. Banking fraud detection, promotional mailing, market
research, supply chain management, tracking stolen credit cards, and
antecedent check by credit companies are some of these applications. The
potential market for these products is estimated to be several hundred
million dollars. Indian software companies are aware of its potential and
have ventured into writing some useful software applications. The software
is not infallible, but that should not detract from its merits, which
essentially lies in integration, automation and embedded security. The technology can bring to focus artificial intelligence and virtual reality to search large data repositories, identify events of interest and compare templates. Elsewhere, much work has been done to acquire this capability. The rub lies in matching wits, in which the terrorists have an edge. Making events appear unrelated, random and seemingly innocuous is their strong point. Archived information can help in timely detection by piecing together the pattern, and sounding the alert based on past acts of terror. It could thereby preempt attack on a government or commercial facility. |