Information Technology Reference
In-Depth Information
In recent years Web sites have begun using another kind of cookie called a
flash
cookie
, which is a file placed on your computer's hard drive by a Web server running the
Adobe Flash Player. Two attributes of flash cookies have raised privacy concerns. First, a
flash cookie can hold 25 times as much information as a browser cookie. Second, flash
cookies are not controlled by the browser's privacy controls. Some Web sites take advan-
tage of this loophole and use flash cookies as a way of backing up ordinary cookies. That
way, if you delete the browser cookie associated with a Web site, it can be “respawned”
from the flash cookie. A survey by researchers at the University of California, Berkeley,
revealed that more than half of the 100 most popular Web sites use flash cookies, but
only four of them mention flash cookies in their privacy policies [47].
In the previous section we surveyed a few of the many ways that companies collect
information on people's daily activities. In this section we look at how this information
has itself become a commodity that companies buy and sell in order to provide more
personalized services to their existing customers and to target potential customers more
accurately.
Before you use a grocery store's loyalty card, you have to spend some time filling out an
application that asks for a lot of personal information, such as your name, address, and
phone number. After the store has processed your application, using your loyalty card is
easy. You just swipe your card or type in your phone number, and the register recognizes
you as a customer and gives you the appropriate discounts on your food purchases. At
the same time, information about your purchases is entered into a database.
A record in a database records a single transaction, such as a particular item you
purchased at the grocery store. A database record is like a single snapshot of a person. It
tells you something about the person, but in isolation its value is limited.
Data mining
is the process of searching through many records in one or more databases looking
for patterns or relationships. Data mining is a way to generate new information by
combining facts found in multiple transactions, and it can also be a way to predict future
events. By drawing upon large numbers of records, data mining allows an organization
to build an accurate profile of an individual from a myriad of snapshots.
Google's personalized search and collaborative filtering are two examples of how
companies are using data mining to create more personal relationships with their cus-
tomers [48].
GOOGLE'S PERSONALIZED SEARCH
Google keeps track of your search queries and the Web pages you have clicked. When you
type in a new query, it can use this information to infer what you are interested in and
return pages more likely to be what you are seeking. For example, the word “bass” has
multiple meanings, but if you have a history of queries and page clicks related to fishing,
but not music, that can help the search engine return the most appropriate pages.
