The ability to modify and differentiate an item or webpage based on implicit data.
Many Business-to-consumer websites (like Amazon) use personalization to
offer suggested future purchases based on previous items bought.

The concept of personalization first became popular through the Internet and Intranets
but later was adapted to almost anything. In common use, it is synonymous with
customization.

Personalization VS Customization
Online, personalization tends to be more automatic than customization.
Rather than a user modifying the colors or layout of a web page, pages that
feature personalization get data based on the habits of the user.

This does not mean that there is absolutely no effect from the user's choices.
When a user creates a profile on a website, he or she may end up choosing
a list of interests which the website would use. This is also a form of
personalization.

Implicit personalization focuses solely on information filled out. Explicit personalization
are based on the user's actions. Finally, many websites use both kinds of information
for hybrid personalization.

Methods of Personalization
Personalization does not have defined categories. Looking through the internet,
one can find many different explanations of personalization, each with their own
take on how personalization is divided. Wikipedia defines the division as:

    Profile Based
    Collaboration Based
    Behavior Based

Profile based personalization looks at a user's profile, such as gender or age
range, and gives results based on that data.

Collaboration based personalization use large amounts of data. Amazon.com
is an example of collaboration. It uses data from items viewed, or purchased,
and combines all the data in order to give a result.

Behavior based personalization attempts to track users as they progress
through websites in order to model their behavior. If a website were combined
with an e-mail system, it could attempt to figure out what a user may have
wanted, but changed his/her mind about. After the user leaves the site without
having bought the item, the site would e-mail a special offer or coupon for
that item.

Technologies of Personalization
One very popular way to personalize website is through Cookies. The saved
data is read by the website and the recommendations are offered. The
automation of such is a system can also be viewed as a weakness,
however. Since the website reads all the data, there isn't a filter for the
user's tastes. If a person were to buy a gift for a friend, the website
would automatically assume that the gift applies to the person's own
tastes.

There are also many different algorithms designed to make personalization
more accurate. This includes the methods used to filter a website's contents
based on the tastes of the user (see collaborative filtering.)

Websites are not the only sources of personalization located online. Web
mailing lists and search engines also use personalization. E-mail lists
use personalization to send advertisements based on a person's apparent
preferences. Google is also using personalization to better target the
search results for the user. 

Disadvantages of Personalization
Personalization can offer many benefits, and save a lot of time for users.
However, it isn't something that is always considered helpful or useful.

Some people enjoy being anonymous online. The internet has many
dangers from having one's information released or found out. Personalization
uses a lot of personal information that some may not want to be known.

 Not everyone trusts the websites they go to or always wants continuing products or
services from them. It may just be a one-time visit or just to find out more information.
However, many "free" offers require a lot of forms to be filled out (like Hotmail)
visit to check out a product. Many times, the users will fill out false information anyway. 


 
 
Wikipedia Entry
Good examples of personalization
Arguement about personalization (though possibly outdated)