Posts Tagged ‘Pagerank’

Just completd reading of The Google storypagerank explained in hemasunder by David A.Vise

here some notes i have taken from that book about page rank

“I’m Sergey Brin and that’s Larry over there. I’m gonna jump right into it here,” Brin began in his typically informal style.
“What we’ve tried to do with our search engine is not give outrageous answers based on the count of words in the text,”
“Let me tell you what the challenges are of a search engine,” Brin said. “You have to index the entire Web. Quite a bit of data. We’ll show you how to manage it. We’ll tell you what it is we do to generate better search results, what we expect to get into three years down the road, and the social issues also.”
The Google search engine took more factors into account than any other search engine on the market. It didn’t just count words or count links and deliver results. It combined information about links and deliver results. It combined Information about links and words with other variables, in new and interesting ways that produced better search results. For example, Brin said, it mattered whether words or phrases on Web pages were close together or far apart, what their font size was, whether they were capitalized or in lowercase type.
“Every time you create a link,” Page told the hushed audience, “you’ve created a citation. But if you just try to count citations on the Web, which is what a lot of search engines do, you run into problems. The Web isn’t like scientific Literature, because anybody can produce Web pages.
“One way to think of PageRank,” he explained, “is it’s a usage model. You have a random [Web] surfer. It’s kind of like a monkey, just somebody who sits around and clicks links all day and doesn’t have any intelligence. You could argue that this kind of approximates people’s behavior on the Web.” Page paused, the audience chuckled, and he went on.
“PageRank is basically saying, if somebody points to you, you get some fraction of the importance that they have. Let’s say somebody really important points to you. That’s worth more than someone who has a random Web page. For example, If Yahoo points to you from their homepage, that’s a big deal. If you just have one link on the Yahoo homepage, that’s pretty good. Either you had to pay someone a lot of money, or your page is pretty good. If have a link on my homepage, nobody would pretty much care.” Page then explained how he derived his recipe for producing ranked search results. “We’ve assigned numbers to those pages to correspond roughly. “We’ve assigned numbers to those pages to correspond roughly to their importance. The page’s ranking is the sum of all things pointing to it.”
“We crawl the web, which means we go out and download the entire Web. We download roughly 100 pages per second,”
“Whenever you query with more than one word, we’re looking at the distance between words [on a Web page],”

that all

Please read the book to get more information about the google and pagerank

Advertisements