Publications

BY VIRGIL LARSON, Omaha World-Herald, October 27, 2007

The title of the first paper listed on the conference schedule says this is no Mary Kay sales meeting: "Succinct Matrix Approximation and Efficient k-NN Classificaton."

Just to get to Omaha to present papers with titles like that was a challenge. Fewer than 20 percent of the papers submitted were accepted by a judging committee.

The meeting is the seventh annual IEEE International Conference on Data Mining. Last year's conference was in Hong Kong. Next year's will be in Pisa, Italy.

"This is a platform to exchange research findings," said Yong Shi, a University of Nebraska at Omaha professor who is a co-chairman of the convention.

"Everything is new this year," Shi said regarding the difficult judging that rejected 80 percent of papers submitted.

In terms of economic impact for Omaha, this is far from either the College World Series or Berkshire Hathaway's annual meeting. Just 320 people will be meeting Sunday through Wednesday at the Embassy Suites Hotel in the Old Market.

Seventy percent of those coming are academics, the rest practitioners, Shi said. They will talk about using computers to dig deeper into data, and how to slice and dice information to make it more useful in a variety of fields.

Data mining is what allows a business to target its advertising more closely to those customers likely to be interested in its products, or a credit card company to spot ID theft or fraud by identifying card uses that are out of a customer's pattern.

That's called "business intelligence" these days.

"The driving force of business intelligence is data mining," said Shi, who does his academic work at the Peter Kiewit Information Science, Technology and Engineering Institute.

Shi said there are "huge" applications, less noticed by most people, in medical research, for example in HIV research. Gregory Piatetsky-Shapiro, a member of the conference's steering committee, said he had done work on brain tumor research through data mining.

Another potential field, Piatetsky-Shapiro said, is personalizing medicine by, for example, creating a prescription designed to meet an individual patient's genetic makeup. Studies show people with similar diseases and similar treatments can get different outcomes because of their "genetic signatures," he said.

Still, the hot fields are business related, Piatetsky-Shapiro said. Companies are digging into the information of people who have laid out their lives on online social sites, he said.

"Facebook and MySpace both look at ways to present more targeted ads."

Asked if he could point to a success that came about as a result of a conference like this, Piatetsky-Shapiro had a one-word answer: Google.

Google's founders, Sergey Brin and Larry Page, attended such conferences in the 1990s, he said.

"I wouldn't claim that just because of attending conferences they founded Google," he said, but Google resulted from developing the results of academic research.

So why do 320 people from 47 countries converge on Omaha when, it would seem, they could gather online and exchange their research and comments?

"Some want to visit beautiful Omaha, Nebraska," said Piatetsky-Shapiro, who runs a consulting service on data mining in Brookline, Mass.

People attend to see each other, he said.

"To visit interesting places, network, see friends you haven't seen in a year. . . . Even nerds like us like to go to these conferences."

Read more.