Wednesday, April 22, 2015
Big data: 'We see you.'
Jing Shyr is not afraid of big data.
In fact, the IBM statistician spends her days trying to find better ways to hoover up the personal information that you spread around the Internet like a drunken sailor, and use that information, well, to better know what you're going to do and want next.
"Every time you try to explain predictive analysis, people think it's 'The Matrix,'' said Shyr, who works for IBM's Analytics division in Chicago. She said it's closer to weather forecasting. Is there a chance of rain? What does the data suggest? Should I bring an umbrella? That's predictive analysis.
I wanted to talk to Shyr—pronounced "sheer"—because the description of her field was so obscure, to me. In the announcement last week naming her an IBM Fellow—the company's top technical innovation honor—she is identified as "a widely recognized leader in the field of predictive analytics. Her methodology for end-to-end automation of statistical analysis became the core of IBM SPSS Analytic Catalyst, a tool that enables non-specialist users to derive forward-looking insights from data."
My hunch was, that means you can fire your middle managers and run your sales stats through an IBM program that tells you whether you'll need to stock more mittens in February. But I wanted to find out.
"We watch web behavior of all kinds," she said, "listening what you are doing. What do you say on the blog? What do you tweet? Who do you follow? We take the behavior data, we understand you, we see you."
And to think some find that ominous. She doesn't.
"Technology is helping humans to articulate their thoughts, to make humans much smarter," she said.
Two keys I keep in mind when approaching technology. First, every single device we use, bar none, was seen by some as the opening gong of doom when it first appeared. Gas lights were an offense to the night. Cars were fiendish. Computers would "take us over."
Oh wait, that sort of happened, didn't it? Yet we don't mind. Which leads to the second key: people change along with technology. If the thought of billboards reflecting what TV show you watched last night seems intrusive, so did installing a telephone in your home, once.
"If you really know how to connect the data, to give you a more clear view," Shyr said. "You get a much more focused and effective type of advertising. Now its still a little bit like junk mail."
She worries herself about the online world.
"As an individual, I am very very afraid to let other people know who I am," she said. " I'm very afraid someone will steal my identity. I'm anxious about that. .... I am very troubled by all types of advertising. It becomes very annoying. "
I told her a story about Amazon, which, like many writers, I tend to despise. But I ordered three books last Friday, they came--miribile dictu--on Sunday, two days later, via the US Postal Service. But only two books. I hurried online, thinking I would have to wage a private war against Amazon to get credit for the missing third book. The information was all there. The shipment split, the tardy book on its way. I felt more positive about Amazon.
"The future is about relationship between vendors and customers," she said. "You give your data, and that is trust. The company knows more about you, they can do the right thing about you. The company holds data and really knows you. How do they know you? They know the data: how often you shop, what is your behavior. Your likes and dislikes, Everything you do and don't do, that 's giving them information."
So be afraid if you like. But this is happening, and some people are thrilled.
"I am very very excited by this data," Shyr said. "Once you feel someone is taking the data, and it is for your own sake, predicting what it means, doing the things you want, it's magic. That's what I see. Predictive analysis can help society. I always explain to people: machines don't get tired. You just need to get them data. Then the machine helps humans become more focused and more efficient."