You are currently browsing the monthly archive for August 2007.

In the coming days, it may be more difficult for me because of the lack of French language skills. Lausanne and Geneva are on the French speaking cantons of Switzerland and that is where most of my day to day life would revolve in the coming months. The little German learned during the Synopsys Aachen days in Germany wouldn’t be of much help, because Zürich is quite far away from Lausanne. I do not see too much of trouble within EPFL, but outside, it is going to be fun game, communicating. I wish, I did learn some French from the Conexant gang in New Jersey. Host of folks were all there to help me out, but I was perhaps ignorant and partly they were trying to improve their English. The urge for me to learn a new language was far too little compared to their keenness to expertise in English. What a wasted opportunity then! There was a formal way to join the Alliance program in Bangalore, but that too was given a cold shoulder. Now, all left to the learn all by practical method there in Lausanne then. For a moment, I had the initial French learning program at EPFL in mind, but the work commitment and other logistics wouldn’t have helped to do that. I shouldnt say that I am worried too much, but the mere fact that, the communication mode would have been set, if only I knew little French before I reached there!

Now, I am relying mainly on Google translate to arrange an accommodation there in Lausanne. Google translate is pretty handy so far, to figure out the difference between studio and apartment. It is a lot fun. I try to derive the meaning of some individual French words, in the process. These are definitely the first few lessons. My friends Vivek and Zarina apparently learned French there in Montpelier, France. They claim to speak decent French. If they could, there is no reason, I cant. May be not all hopes are lost. Let us wait and watch. In few months down the lane, I should be able to blog on in French.

Some sort of history making on yesterday at the Oval cricket ground in London[1]. India, after a gap of 21 long years registered a test series win in England against England of course! The time 21 itself is a far stretched number, by all means because, only 4 times (1990, 1996, 2002 and 2007) these two teams played together in England during this period. On equivalent terms, it is a win after 3 previous attempts went against the wish list.


One of the practical applications of linear algebra is the success of Google. If a math teacher gives out this response to students, that would certainly produce many Ahh Huuu…, but truth indeed is there in the statement. When it comes to Google and its rocketing success story, some may wonder whether mathematics has so much of practical applications in it! Kurt Bryan and Tanya Leise aptly mentioned [1] that this eigenvector is worth more than $5,000,000,000! Aha, isn’t this the richest eigenvector that you ever come across?

One of the core technique behind Google’s multi billion dollar success story is the page rank algorithm, developed by its co founders Larry page and Sergey Brin, while they were in Stanford[3]. Let us put the statement mathematically or rather linear algebraic: It is essentially ranking web pages according to an eigenvector of a weighted link matrix. So, Google search has its thrust based on solving this eigenvector computing! Computing eignevalues and eigenvector, are sole linear algebra problems. The deal is quite big though. Let us talk a little bit deep about this problem.

Google’s website [2] has only modest thing to say about this fantastic algorithm:

PageRank relies on the uniquely democratic nature of the web by using its vast link structure as an indicator of an individual page’s value. In essence, Google interprets a link from page A to page B as a vote, by page A, for page B. But, Google looks at considerably more than the sheer volume of votes, or links a page receives; for example, it also analyzes the page that casts the vote. Votes cast by pages that are themselves “important” weigh more heavily and help to make other pages “important.” Using these and other factors, Google provides its views on pages’ relative importance.

Of course, important pages mean nothing to you if they don’t match your query. So, Google combines PageRank with sophisticated text-matching techniques to find pages that are both important and relevant to your search. Google goes far beyond the number of times a term appears on a page and examines dozens of aspects of the page’s content (and the content of the pages linking to it) to determine if it’s a good match for your query.

A basic listing of the pagerank is available here at Here is the summary extracted from there.

  • PageRank assigns a rank or score to every search result. The higher the page’s score, the further up the search results list it will appear.
  • Scores are partially determined by the number of other Web pages that link to the target page. Each link is counted as a vote for the target. The logic behind this is that pages with high quality content will be linked to more often than mediocre pages.
  • Not all votes are equal. Votes from a high-ranking Web page count more than votes from low-ranking sites. You can’t really boost one Web page’s rank by making a bunch of empty Web sites linking back to the target page.
  • The more links a Web page sends out, the more diluted its voting power becomes. In other words, if a high-ranking page links to hundreds of other pages, each individual vote won’t count as much as it would if the page only linked to a few sites.
  • Other factors that might affect scoring include the how long the site has been around, the strength of the domain name, how and where the keywords appear on the site and the age of the links going to and from the site. Google tends to place more value on sites that have been around for a while.
  • Some people claim that Google uses a group of human testers to evaluate search returns, manually sorting through results to hand pick the best links. Google denies this and says that while it does employ a network of people to test updated search formulas, it doesn’t rely on human beings to sort and rank search results.

[1] Kurt Bryan, Tanya Leise, The $25,000,000,000 eigenvector. The linear algebra behind Google. SIAM Review, 48 (3), 569-81. 2006


This is the photo taken in November 2006, when Ajay had come down to Bangalore. We had gone to the food court (Myself, maya, nivedita, ajay, soni and ajay’s mom)

Ajay, Soni and myself facing the camera:-left to right!)


August 2007
« Jul   Sep »

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 86 other followers


%d bloggers like this: