Off-Topic :
Fairfax Underground
Welcome to Fairfax Underground, a project site designed to improve communication among residents of Fairfax County, VA. Feel free to post anything Northern Virginia residents would find interesting.
Welcome to Fairfax Underground, a project site designed to improve communication among residents of Fairfax County, VA. Feel free to post anything Northern Virginia residents would find interesting.
Data Mining Social Media and Search History for Loan Applicants
Posted by:
eesh
()
Date: October 12, 2013 10:44AM
http://arstechnica.com/business/2013/10/denied-for-that-loan-soon-you-may-thank-online-data-collection/
http://www.scientificamerican.com/article.cfm?id=rich-see-different-internet-than-the-poor
Per an article Crawford cited from Scientific American, redlining specifically prohibits companies from differing their offers to potential customers based on what they know about them. For instance, they can’t offer a loan at six percent interest to someone who lives in Tonyville, CA and then turn around and offer to same loan at 12 percent interest to someone who lives in Boondoggle, TX.
In the old days, if both of those customers saw the six percent offer on a billboard and walked in the bank to apply, the bank couldn’t legally offer the customer from the bad neighborhood the higher interest rate based on where they live. But with online targeting, the bank can make sure the bad-neighborhood customer never sees the offer, period, avoiding the perceived “risk” altogether.
Crawford cited some examples that fall to onto the barely-regulated end of the spectrum: a WebMD search about breast cancer plus a book you buy from Amazon about cancer survival, mixed with a few tablespoons of anonymization and a sprinkling of cross-referenced geographical/demographic data to negate that anonymization, yields you unable to get health insurance or approved for a loan. Why? Data indicates you might possibly be dead soon. It doesn’t matter so much that the breast cancer search and book were out of curiosity and generosity, respectively, for your dying great aunt who has never touched the Internet.
Edited 1 time(s). Last edit at 10/12/2013 10:44AM by eesh.
http://www.scientificamerican.com/article.cfm?id=rich-see-different-internet-than-the-poor
Per an article Crawford cited from Scientific American, redlining specifically prohibits companies from differing their offers to potential customers based on what they know about them. For instance, they can’t offer a loan at six percent interest to someone who lives in Tonyville, CA and then turn around and offer to same loan at 12 percent interest to someone who lives in Boondoggle, TX.
In the old days, if both of those customers saw the six percent offer on a billboard and walked in the bank to apply, the bank couldn’t legally offer the customer from the bad neighborhood the higher interest rate based on where they live. But with online targeting, the bank can make sure the bad-neighborhood customer never sees the offer, period, avoiding the perceived “risk” altogether.
Crawford cited some examples that fall to onto the barely-regulated end of the spectrum: a WebMD search about breast cancer plus a book you buy from Amazon about cancer survival, mixed with a few tablespoons of anonymization and a sprinkling of cross-referenced geographical/demographic data to negate that anonymization, yields you unable to get health insurance or approved for a loan. Why? Data indicates you might possibly be dead soon. It doesn’t matter so much that the breast cancer search and book were out of curiosity and generosity, respectively, for your dying great aunt who has never touched the Internet.
Edited 1 time(s). Last edit at 10/12/2013 10:44AM by eesh.