How does GOOGLE work ?

You must have come across GOOGLE at some point of time or other. Most probably you have also used it while surfing the internet. Have you ever wondered about it's functions ? Ever thought what technology does it use that it is able to deliver search results so fast and so accurately ?

On average an ordinary search for a file in our computers take more time than that taken by Google to search for websites on a particular topic ; and Google has a lot more number of records than our computer's number of files.

Google runs on a unique combination of advanced hardware and software. The speed that we experience can be attributed in part to the efficiency of their unique search algorithm and partly to the thousands of low cost PC's that they have networked together to create a super fast search engine.

In the heart of Google, there is a unique Page Rank system, which ranks pages. The main method was developed by the two founders of Google, Larry Page and Sergey Brin at Stanford University. This method has been later developed and modified by dozens of engineers.

This Page Rank method relies on the unique democratic nature of the World Wide Web by using its huge link structure. In this method, a link from page A to page B counts as a vote for page B. The value of this vote depends on the page which contains the link to page B. If that page is an “important” page (this is determined by checking it's page ranking), then the value of the vote is high. In this way it continues to sum up all the points thus earned. Google records the time when they last checked a particular page and after a certain period of time checks that page again.

Google uses complex programs called web crawlers, which are capable of visiting new pages by following links from one page to another page. These web crawlers, apart from reporting the links also does a sophisticated keyword search and stores the result. Words that appear in headings and names of pages are given priority. Also the number of occurrences is checked. They even match these words with that of the page containing the link to this page !

Based on all these information, Google has created a database with keywords. When we search for a topic, websites containing the topic are shown. The pages with a higher page rank are shown first.

While the link counting part was easy, the most difficult part is to search for keywords in a page and then classify it accordingly. Google even has the ability to throw up search results for something as bizarre as : banana+speaker !! How do they retrieve the data so fast is a point to ponder. Till date, Google has not given any plausible answer to this query.

In Google, image searches, are carried out mainly on the basis of the name of the images and later matching them with the query string. Similarly research papers, articles, white papers can also be seared for. Google gives the users options to view cached pages (to view previously stored pages of the sites that they have thrown up in the search).

But Google has also got it's own limitations. It searches images based on names of images! It cannot perform context specific search, i.e. if I search on “bird”, it can throw up pages containing information on birds or autobiography of Dicky Bird!

However, till date Google is one of the most reliable search engines around. Though search engines like Yahoo! exist, which categorizes sites manually, they are slow and often do not contain recent information. Researchers of Google claim that they are on their way to make a context specific search engine, but according to the pundits, that day is far off.