Future of Internet----Web Personalization

I like Chinese food but my sister is a big fan of Italian food. When we were kids we used to fight before going to any restaurant. She always preferred to eat pasta and I used to force her to eat manchurian. In the end we both were forced to eat in an Indian restaurant of my parents’ choice or any one of us had to compromise as we were not able to make any unanimous decision. But nowadays my father has found a solution. We go to food court in a mall where we get all stuffs Chinese, Italian and Indian at one place without compromising ourselves. These days no one wants to compromise and one tries to customize everything according to ones taste like whether it is ones room or ones bike , So why not our social networking site or why not my search engine gives me the result which I like or want without entering any term. These questions cause Web personalization to come into the market. Web personalization means providing information to different tastes of people at one place like food court. It is very simple and cheap to develop a website; any company can do it but to make the difference Web personalization comes into existence. If you search Wikipedia for a term “personalization” you will get following definition "Personalization involves using technology to accommodate the differences between individuals". Each individual has different tastes and needs, so personalization means organizing all the information in more logical way and in such a way that an user can find things easily and in the way he wants.

Now we shall discuss how personalization is used these days in websites:

• When you open google news it will give you only your regional news (Ex. India for Indian Users.) You have nowhere mentioned which country you belong to.
• When you go to amazon.com it gives you suggestion what product you may want to buy?
• Job portals offer you jobs from your city.
• You can select your own orkut theme for your profile.
• MyYahoo in which you design your own dashboard.
• You get friend suggestions in Facebook.
• You get your recommended videos in YouTube.

Have you ever think how these Facebook guys know my friends?

How come YouTube can give you suggestion of your favorite video without asking you anything?

Before answering these questions, I will tell you different types of personalization. If you search on web you will get various divisions of personalization, but, when I was doing my collage project work on web personalization with my friend Paul, we concluded that it could be divided into 2 simple types, Automated personalization and Manual personalization. Manual personalization means the user will directly tell what he wants. For Example user will manually tell what theme he wants to use in orkut. And automatic personalization is without directly asking from the user the web site will provide him information, like friend suggestion in Facebook. Implementing manual personalization is simple as we get direct input from the user. But complexity of automatic personalization differs, as we have to implement different algorithms to produce desired output. I will give you an example. Have you seen recommended videos in you tube. This section will provide you the videos which you may like. So, for generating this output an algorithm is written which performs the following tasks:

• It will consider your last searches on the website. As it is highly probable that you will search things which you like.
• It will scan your favorite videos which you have already added and then try to find out similar videos, like if in your favorite videos there are many songs of Ricky Martin then there is high probability that you may like his other songs too.
• Now as YouTube and Orkut belong to same company, YouTube can access Orkut’s data. In orkut we have profiles in which we have favorite songs and favorite movies sections. So from these sections we can suggest what a user likes.

There can be many more criteria, but these are all I can think. Now using this information algorithm will calculate user’s favorite videos.

This automatic personalization algorithm is little bit complex. I will give you one more example. I have observed that in YahooMail people of age between 20-30 get matrimonial ads on the website. As it is highly probable that they want to find their partners and they will click that ad. Similarly in Facebook we get friend suggestion. This can be done by finding common friends. It is highly probable that if you have more number of common friends with one person then the person may be your friend. It can also check that if you are from same college or school then the person can be your friend. There may be many more criteria which they (Facebook/Youtube) are using.

This is how they implement personalization on their sites. It is something big these days. This is the reason why web 3.0 is about personalization.