数据挖掘：是福是祸

 

"Pizza Palace, may I take your order?" a woman's voice echoes from a video clip that plays in the Giovanini Commons of the University of Notre Dame's Mendoza College of Business.

"Is this Mr. Kelly?" the woman continues.

"Yes," a man replies.

"Thank you for calling again, sir," the woman responds. "I show your national identification number as 6102049998-45-54610 — is that correct?"

"Uh, yes..." the man answers.

"Thank you Mr. Kelly," the woman says. "I see you live at 736 Montrose Court, but you're calling from your cell phone. Are you at home?"

"Uh," the man replies. "I'm just leaving work, but I'm..."

"Oh, we can deliver to Bob's Auto Supply," the woman says. "That's at 175 Lincoln Avenue, yes?"

"No!" the man says, frustrated. "I'm on my way home! How do you know all this stuff?"

"We just got wired into the system, sir," the woman says.

The video clip — a cautionary dramatization circulated by the American Civil Liberties Union — was enough to draw a few laughs from the audience during the final event of the 12th annual Ethics Week.

"What they're talking about here is a situation where information has been (collected) from a lot of different sources," said Barry Keating, professor of finance at Notre Dame, during the session, which focused on data mining. "And it has been put together and used by a single company."

But take a deeper look at the concept of data mining, and the ethical dilemmas surrounding it can be pretty serious.

"Data mining can be used for the good of mankind," said Keating. "But it can also be used perhaps to the detriment of mankind."

What is data mining?

In a nutshell, data mining can be defined as extracting useful information from large data sets.

Data mining, Keating said, plays a key role in today's society since it's used in fraud detection all the time.

This includes detection of credit card fraud, money laundering, phone fraud, securities fraud and medical fraud.

Data mining, Keating added, can be applied in a variety of fields, including scientific areas of astronomy and drug discovery, governmental areas of law enforcement and profiling tax cheaters, and business areas of advertising, manufacturing and customer relationship management.

Keating cited specific examples, such as in e-commerce, where a person buys a book at amazon.com and receives recommendations of other books he or she is likely to buy.

But with the beneficial uses of data mining come some key concerns.

Ethical issues

When it comes to data mining, ethical issues may be more accurately described as problems in data security and privacy preservation, Keating said.

In fact, the most fundamental ethical issue deals with the basic storage and retrieval of personal data, he said.

"Where did you get the data?" Keating said as an example. "Whose data is it? Did somebody allow you to use their own data?"

For Keating, a big test of ethics and data mining is probably going to surround a search engine people use every day — Google.

Take, for example, how Google unveiled a new service last year that will allow you to store and access your medical records on the Web.

And just this month, Google released software that allows users of mobile phones and other wireless devices to automatically share their whereabouts with family and friends. This allows users in 27 countries to be able to broadcast their location to others constantly, using Google Latitude.

Through sheer speed of collection, Google will test the limits of what our society can tolerate, Keating said.

"Google really is going to test the limits of the ethical dimensions of data mining," he said.

The question is, do we know what we're getting ourselves into by giving up so much information about ourselves?

"I wonder whether everybody knows what they're willingly giving up," Keating said. "I wonder if I know what I'm willingly giving up. Am I giving up too much?"