giovedì 12 aprile 2012

OT: l'analisi quantitativa di una vita (al computer)

Il New York Times dedica un articolo all'analisi quantitativa di 23 anni di posta elettronica che potete leggere sul blog di Stephen Wolfram.

One day I’m sure everyone will routinely collect all sorts of data about themselves. But because I’ve been interested in data for a very long time, I started doing this long ago. I actually assumed lots of other people were doing it too, but apparently they were not. And so now I have what is probably one of the world’s largest collections of personal data.
Wolfram è un tipo un po' particolare: oltre ad aver conservato i messaggi di posta elettronica ricevuti e spediti negli ultimi 23 anni ha anche registrato tutti i tasti che ha battuto sulla tastiera del suo computer...
 For many years, I’ve captured every keystroke I’ve typed—now more than 100 million of them
e tutte le telefonate che ha fatto (negli ultimi anni ha trascorso in media quasi 5 ore al giorno al telefono, non lo invidio proprio...)

Secondo Wolfram e il NYTimes questo tipo di registrazioni 
... may someday end up serving as a kind of personal historian, as well as a potential coach for improving work habits and productivity. The data could also be a treasure trove for people writing their autobiographies, or for biographers entrusted with the information.
Dr. Wolfram has also scanned 230,000 pages of paper documents and, when possible, fed them through an optical character reader. He has at the ready his medical test data, complete genome, GPS location tracks, room-by-room motion sensor data and much more — all possible fodder for future analysis.
People, sometimes known as self-quants, have been hard at work in the medical arena. One is Larry Smarr, a computer science professor at the University of California, San Diego, who also directs the California Institute for Telecommunications and Information Technology in La Jolla.
Dr. Smarr wears one wireless sensor to monitor the calories he burns and another to see how well he sleeps. He is keeping track of the bacteria in his body, about 100 different variables in his blood and many other fine points in his biochemistry.
After examining the data, he makes changes to improve his health. (So far, he’s lost weight and gained hours of deep sleep, he says.)
“There’s so much information you can track,” Dr. Smarr says. “And the cost of measuring and analyzing the data about ourselves just keeps on decreasing.”
Non sono sicuro di cosa voglia dire, e ancor meno sono sicuro dell'uso che un giorno sarà possibile fare dell'ingente mole di dati personali che più o meno inconsapevolmente ogni giorno lasciamo dietro di noi. Tuttavia non sarei così certo di quanto scrive, nello stesso numero del NYTimes, uno psicologo, rispondendo alla domanda se gli algoritmi possano o meno calcolare in modo attendibile l'affinità tra due persone: 
It may not be a problem that software can solve on its own, said Eli Finkel, a professor of social psychology at Northwestern University. “Technology is not the way to figure out who is compatible and will never be,” he said. “At the end of the day, the human algorithm — neural tissue in our cranium called a brain — has evolved over a long period of time to size up people efficiently. On a blind date, a person arrives and in that instant I can say I’m glad I did this or regret it.”
Professor Finkel, along with several other researchers, published a study this year raising doubts about the idea that a personality test or algorithm of the kind popularized on eHarmony, can help you meet a potential mate.
Sites that say algorithms can help you find your soul mate “are probably spitting in the wind,” said Harry Reis, a professor of psychology at the University of Rochester and a co-author of the algorithm paper, who has written upwards of 120 papers on online dating.
EHarmony counters that the algorithms it uses do work, citing research it conductedinvestigating the satisfaction of couples who met through the site, and their divorce rate.
The system that eHarmony has built is “based on years of empirical and clinical research on married couples,” said Becky Teraoka, an eHarmony spokeswoman. They include “aspects of personality, values and interest, and how pairs match on them, that are most predictive of relationship satisfaction.”
While Professors Finkel and Reis question the value of algorithms, they do say that online dating is useful because it can broaden the pool of people you come across on a regular basis.

Nessun commento: