Innovation that matters: Jeff Jonas connects the invisible dots

He's developed software that not only connects the dots, but shows users dots they never knew existed -- whether to find terrorists or reunited separated families

When you think of research breakthroughs from IBM, Big Blue's famous Watson and Almaden labs and their corps of highly educated researchers and engineers come to mind, of course. But how often do you hear of an IBM "distinguished engineer" whose formal education ended with a high school equivalency diploma?

That man is Jeff Jonas, chief scientist of the company's Entity Analytic Solutions. His research, which started when he was just 16 and living in his car, has led to an innovative new take on identity and business intelligence. Systems developed with Jonas's technology have been used to defeat card counters in Las Vegas, analyze connections among terrorists, and help reunite families separated by Hurricane Katrina.

[ Stay ahead of advances in technology with InfoWorld's InfoWorld Daily newsletter. ]

Much of his thinking can be summed up in a phrase he likes to sprinkle into conversations: "The data is the question."

Data Explosion iGuide

Here's how he explains that aphorism: "Next-generation 'smart' information management systems will not rely on users dreaming up smart questions to ask computers; rather, they will automatically determine if new observations reveal something of sufficient interest to warrant some reaction; for example, sending an automatic notification to a user or a system about an opportunity or risk," he and co-author Lisa Sokol wrote in an essay last year.

See the dots, find the felon
Imagine that investigators at a large bank are aware that fraud has been perpetrated and the culprit probably had an inside accomplice. Then suppose a human resources person changes his phone number. Ordinarily, there would be no reason to see a connection between those events.

Now suppose that the HR person's new phone number also appeared in data collected during the investigation. The chances of someone, or a typical information system, making a connection are remote. Indeed, in all probability no one has even asked that question. But software developed by Jonas before his company was purchased by IBM, and now marketed as IBM's Identity Insight, would make that connection in seconds and alert the proper person without being asked, he says.

Although this sounds a bit like a data warehouse, it's not. "It's a service to a warehouse," says Jonas.

What happens in Vegas
Jonas came into his own as a technologist in Las Vegas, and not surprisingly the work of his company, Systems Research Development, focused on the gambling industry. The company developed a system called NORA, or nonobvious relationship awareness.

"It ran on a single server, accepted data feeds from numerous enterprise information systems, and built a model of identities and relationships between identities (such as shared addresses or phone numbers) in real time. If a new identity matched or related to another identity in a manner that warranted human scrutiny (based on basic rules, such as good guy connected to very bad guy), the system would immediately generate an intelligence alert," Jonas wrote for an IEEE publication.

[ Discover what's new in business applications with InfoWorld's Technology: Applications newsletter and Killer Apps blog. ]

Leveraging facial recognition, this technology enabled casinos to defeat aggressive card counters, including the "MIT team," the subject of the book "Bringing Down the House." In 2001, SRD caught the attention of In-Q-Tel, the venture capital arm of the CIA. NORA is credited with developing a unique analysis of the connections among the 9/11 terrorists.

Following IBM's acquisition of SRD in 2005, NORA's underlying code base was improved and the technology renamed IBM Identity Resolution and Relationship Resolution. Today, the system is implemented in C++ on top of an industry-standard SQL database with a Web services interface and optional plug-ins (knowledge-based name libraries, postal base files, and so on).

Data integration services transform operational data into prescribed and uniform XML documents. Configuration files specify the deterministic matching rules and probabilistic-emulating thresholds, relationship scoring, and conditions under which to issue intelligence alerts.

If you can't count it
Jonas, now 45, looks a bit like Apple's Steve Jobs: thin, nearly gaunt, with what little hair he has left closely cropped, and wearing black. But he lacks Jobs' nasty edge and can barely contain his enthusiasm for his work. He spends much of his time on the road, evangelizing and appearing at technical conferences; indeed his whole team at IBM is virtual.

At 16, he wrote a word processing program for the Commodore and eventually sold it to the Los Angeles School District. By 17 he had started his own company.

Despite work related to national intelligence, Jonas sees himself as a privacy advocate and has developed software that makes it easier for companies to trade information about customers without compromising the identity of individuals.

He worries that the massive amounts of data collected by GPS-enabled devices could be used to thwart democracy. "A government not so keen on free speech could use such data to see a crowd converging toward a protest site and respond before the swarm takes form -- detected and preempted, this protest never happens. Or worse, it could be used to understand and then undermine any political opponent," he posted last year.

These days, much of his enthusiasm is directed at something that sounds really boring, but isn't: expert counting engines. "If you can't count it, you can't analyze it," Jonas says. Many applications need to know how many people (or things) are represented by a data set. The problem is much more complex than mere data cleaning. Indeed, rather than discarding or simply fixing dirty data, a Jonas counting engine looks at the dirty data in context to help determine how many objects exist.

Next up: technology to serve Web ads "in milliseconds," based on intelligence about the person who is browsing and moving his analytic engines to the cloud. On the personal front, Jonas continues to compete in Iron Man triathlons around the world. He's not your typical researcher, but Jonas is one of the most interesting technologists I've met in some time.

I welcome your comments, tips, and suggestions. Post them here so all our readers can share them, or reach me at bill.snyder@sbcglobal.net.

This article, "Innovation that matters: Jeff Jonas connects the invisible dots," was originally published at InfoWorld.com. Read more of Bill Snyder's Tech's Bottom Line blog at InfoWorld.com.

Join the discussion
Be the first to comment on this article. Our Commenting Policies