I’ve got two weather apps on my phone. They’re both sophisticated examples of turning mounds of data into useful information. And they represent what’s good and bad about the state of Big Data.
The good part is easy. I can get weather info wherever I am as long as I have my phone. That’s great. But just like with Big Data, you have to be aware of how things work and where promises may exceed reality.
You need accurate and timely data
You know the saying, “garbage in, garbage out.” It’s true for big data, just like for little data.
My two apps both use GPS to locate my address so, they say, they can give me an accurate forecast. But they often get the location data wrong. One app occasionally, but not always, puts my house in a different town. The other thinks that my office and the kitchen are at different addresses.
The algorithms have to work all the time
Supposedly both weather programs use similar data and tested algorithms. So why do they often have different forecasts? Beats me. We’re still learning how to do this Big Data thing right.
You want to claim the right level of precision
One of my apps routinely gives me messages like, “rain will stop in 17 minutes.” Really? Delivering a calculation in precise minutes doesn’t make the prediction more accurate if the best estimate is “between 15 and 20 minutes” or “soon.”
You should look outside from time to time
As nifty and techy as these apps are I don’t rely on them when I’m getting ready for my daily walk. I look outside.
Boss’s Bottom Line
We’re all on this learning curve together and it will take us a while to get good at it. In the meantime, make sure you’re feeding the system the right data, use two different calculations to check against each other, be skeptical of precision, and conduct frequent reality checks.
There are no comments yet, why not be the first to leave a comment?