(AMS Bumper Sticker)



Web math-frolic.blogspot.com

Wednesday, February 25, 2015

A Beach Read for the Middle-of-Winter


Like many folks, "big data" at one time interested me, but I became bored with it a year or more ago as it seemed waaaay over-hyped, imprecise, and not living up to promises/claims.  I've avoided most of the popular books out on "big data" waiting for the field to catch up a bit to its vaunted notoriety. Still, having seen a lot of positive reviews of Christian Rudder's "Dataclysm," I pulled it off a library shelf not long ago, and scanned the first 20+ pages... and, he hooked me!
I won't really review the volume, but here are some longer reviews for those wanting such:

http://tinyurl.com/pzycv2y  (by Jordan Ellenberg)
http://mathbabe.org/2014/09/16/christian-rudders-dataclysm/  (a more critical review from Cathy O'Neil)

Christian Rudder was a co-founder of the dating site OKCupid, from which much of the info/data for this book derives (is that a "representative" sample of humanity?). But he's also worn many other hats, and additionally has a mathematics degree from a place called "Harvard" i.e., he's an earnest mathematician/data-analyst. I like that he doesn't seem to take either himself or his subject matter too seriously here, and admits he stayed away from a lot of the deeper possible math to write an engaging book for a general audience.
I view this as more of a recreational book than a terribly serious math, or even human behavior, treatise, and probably am more hesitant than Christian to reach some of the conclusions that are speculatively drawn. But still, plenty of interesting tidbits in a very fun, entertaining package, with a lot of grist for hearty discussions/debates among friends... I can definitely recommend it, if only as, what Steven Strogatz calls it on the back-cover, "a guilty pleasure." Human behavior, is after-all, inherently curious and fascinating, and Rudder does have a wealth of data to draw from, lending us a fun, ultimately hopeful (perhaps Pollyannishly-so?) treatise on the future of "big data."

If it were the right time of year, I would call this book a beach-read for math geeks, but since it's the middle of winter I'll just call it a great weekend read as long as you don't take the conclusions too seriously or uncritically. For now, 'big data' is still in its infancy... perhaps even its gawky, 'terrible twos' stage so-to-speak, but it is advancing rapidly and certainly we need to keep a close eye on it.



No comments: