Analysis & Commentary:

INTERVIEW WITH INDERPAL BHANDARI
by Alan Beck, Editor in Chief

DSstar: It would appear that optimal intelligence can only be gleaned from an ideal balance between HumInt and MachInt. Yet how can such balance be pragmatically achieved? What factors might benchmark whether or not we are on the proper course?

BHANDARI: Absolutely. In fact, optimal intelligence can only be gleaned from a balance between MachInt and not just HumInt but all the other lines of intelligence - SigInt, Elint, etc. My main point is that presently there is no MachInt - analysis is mainly done by analysts assigned to each line of intelligence. Sure, they make use of the latest computers and analytical tools but there is still a manual bottleneck. The bottom line is that people cannot analyze as fast as machines can collect. Consequently, the existing lines of intelligence collect huge amounts of data that then lie largely unanalyzed as this primarily manual analysis cannot keep up with the pace of data collection. Hence, we need MachInt, smart computer programs that can provide cueing or early warning directly to the decision makers in largely automated fashion. With regard to benchmarking, it should be straightforward to determine if a balance is being achieved. The steady-state size of the mountain of data that remain unanalyzed will reduce as should the number of post mortem analyses where it becomes clear after the fact that key items of data had been collected by the intelligence community but missed.

DSstar: Because intelligence is necessarily sensitive and classified, how can safeguards be built in so that decision makers can be guided through effective MachInt utilization without jeopardizing security? In other words, can MachInt technical personnel and integrators be appropriately separated from critical decision-making areas?

BHANDARI: I think staffing safeguards for MachInt will be pretty much what they are for the other lines of intelligence - a comprehensive system of background checks and clearances. The contractors and system integrators who do not have the appropriate clearances will participate only at the level of abstract platforms and sanitized data while the employees with clearances will realize the implementation and educate the decision makers.

DSstar: There is an enormous amount of relevant data available for national defense, and this appears to be growing due to expanded surveillance. If a firm commitment to MachInt were made by the intelligence community, what would be the steps and timeline needed for implementation? What level of funding would be required?

BHANDARI: It is tempting to think that the evolution of MachInt will parallel the creation and growth path of any of the three-letter agencies such as the NSA or the CIA, since these agencies fundamentally started out concentrating on one line of intelligence. For example, the NSA was largely about SigInt and the CIA was largely about HumInt. However, I believe that the benefits of MachInt can accrue faster and cheaper than was true for other lines of intelligence since it can build on data collection and decision making processes already in place. Think of MachInt as a family of smart computers to bridge those processes. It evolves in lock-step with those processes and feeds back to those processes. But it does not have to underwrite those processes since they are developed by other agencies. I believe it is not unreasonable to expect initial results in 6-12 months with annual funding in the range of the high hundreds of millions dollars instead of billions of dollars.

DSstar: Because MachInt represents a comparatively novel approach to intelligence evaluation, would its use inadvertently increase bureaucratic overload? Why or why not?

BHANDARI: Certainly, every time you have a new organization there is the danger of adding to the bureaucracy. However, MachInt will reduce not increase bureaucracy. In order for the smart MachInt computers to provide input directly to decision makers and their staffs, they must have access to and process all underlying data sources that feed into a decision. Therefore, the MachInt infrastructure will introduce a much higher degree of standardization and coordination across different agencies than presently exists, leading to increased efficiency. There will also be cross-fertilization between MachInt advances in automated analysis and the mechanisms that exist in the other intelligence agencies to convert raw data into intelligence. This will lead to a speed-up of those mechanisms.

DSstar: What can those interested in seeing serious deployment of MachInt do to further this end? Are there any other key points we should know?

BHANDARI: We need to educate the powers-that-be that throwing more bodies at the intelligence data problem is not going to work on account of the rate at which that data is increasing. Long-term, the only viable strategy is to create a line of intelligence, MachInt, involving smart computers that provide cueing directly to the decision maker. A key point on MachInt, it is a vision that will be realized over time. In the early going it's implementation will involve more art than science. Fundamental technical challenges must be addressed before fully automated pathways will exist between decision makers and raw data. This re-emphasizes three points touching on ideas alluded to before. First, we need a separate organization to nurture the talent required to do this, much as the creation of the NSA fostered the development of top-notch sigint and crypto talent. Second, MachInt implementations must build on the mechanisms for raw-data-to-intelligence conversion of the other agencies. Third, cross-fertilization between MachInt and other lines of intelligence should be fostered to find the right balance between automatic cueing and manual cueing.

Dr. Inderpal Bhandari, the founder and CEO of Virtual Gold, Inc., is an internationally recognized expert in data mining, the art of using computer programs to discover knowledge from large amounts of data. Under his leadership, Virtual Gold developed its powerful patent-pending VirtualMiner technology that enables business managers to make more accurate decisions by alerting them to hidden patterns in business data that are discovered automatically. While other data mining platforms can find patterns that have statistical merit, VirtualMiner goes much further by discovering hidden patterns that can actually impact the business. Dr. Bhandari has built Virtual Gold into a market leading company with strategic partnerships with companies such as IBM and Dictaphone and blue-chip customers like Merrill Lynch, Chubb & Son, Bank of America, CBS Sportsline, Boston Celtics, Phoenix Suns and other teams of the National Basketball Association and the Professional Golfers Association.

About Virtual Gold Inc

Virtual Gold Inc, www.virtualgold.com, is a leading next generation business intelligence and data mining software company. Former IBM scientist Dr. Inderpal Bhandari, one of the world's leading authorities on data mining and its timely application to real world problems, founded it in 1997. In addition to its MacInt line of products, Virtual Gold provides an integrated family of Web-based data mining software and solutions in the following fields: contact center, customer relationship management, enterprise resource planning, decision support systems, e-business, and sports and entertainment. The company is headquartered in Hartsdale, New York.