Oh! The pages you'll go!
By Jonathan Schindler
Hello World!
The program is done.
We'll be crawling the Web!
We'll be surfing for fun!
We have our starting locale
and everything next.
We'll follow the Web Rules
even robots.txt!
We'll crawl on our own, not sure where we'll know
because the Surfer decides where we'll eventually go.
We'll look down the pages, for info so clear.
Finding what we want, either farther or near.
With our starting locale and everything next,
we'll be surfing the web, following robots.txt
And maybe our local start point
will contain many links,
but on the scale of the Net,
this is only the brink!
It's scary out there,
in the Internet's lair.
With our eyes wide open and map in our hand
we can pick any position, in this virtual land.
Our starting position should be one that such
connects to many others, so it might be in Dutch.
If it's a good page, you can simply connect
to many other hubs, without any regret.
If it's a bad page, you might be at a loss
getting stuck in a place, like a certain albatross.
But fear not young one, a slight memorandum
You have the ability to teleport at random!
Randomly teleporting, to our new node
We are now safe, no need to reload.
Linking and Teleporting, Teleporting and Linking
Surveying the Internet is faster than blinking!
You simply move through that oversized blob
Maybe eventually reaching Caverlee's Blog.
The info you found is stored into data
and with proof of concept, you'll go into beta!
Hello World!
The program is done.
We'll be crawling the Web!
We'll be surfing for fun!
If you record all the links, and what goes to who
you can calculate PageRank, simple as 3 + 2
Just build your Link Matrix, or L for short
And build another matrix: T for Teleport
Our PageRank matrix is really not hard
PageRank equals alpha L plus beta T, right Jaccard?
The alpha is our weight of using our links
and beta is 1 minus alpha, or more simply put, our chance to Blink.
L is built, from percent chance of a shift
so a page with 5 links will result in one-fifth.
But listen now here for the next part is tricky
a page with no links becomes slightly picky.
If the page is no hero,
there is not a link.
The result should be zero
That would really stink!
So you'll teleport instead!
with 1 over N,
where N is the nodes
That's using your head!
The teleport matrix is whatever you'd do,
It could be 0.9 or one over 8,742!
With the equation complete, we're almost right done
It is now time to add the plus one.
The Power Rule comes to enable our bid
take your initial vector and multiply by the grid
Continue this method until they converge
You'll recognize this point when patterns reemerge
The vector returned will contain different ranks
Just return them in order, and now you'll make bank!
Oh! The places you'll go,
the pages you'll see
When surfing the Internet
with Hubs and Authorities.