Return to basic course information.

**Assigned:** Thursday, 27 September 2012

**Due:** Friday, 12 October 2012, 6pm

Students have to submit a **hard-copy / paper copy**
of their report including the details wherever 'To hand in: ' is
present in the project problem statement. You can hand it in at
Thursday's lecture, or at Madan's office hours on Friday 1-3 and
5:30-6.

The only part that should be emailed is the source code in form of zip file. The zip formats are already mentioned in the first post of 'Submission Guidelines' on Piazza. TAs will schedule a 5 minute demo wherein you may be asked specific questions about your implementation.

// P is the set of all pages; |P| = N // S is the set of sink nodes, i.e., pages that have no out links // M(p) is the set of pages that link to page p // L(q) is the number of out-links from page q // d is the PageRank damping/teleportation factor; use d = 0.85 as is typical foreach page p in P PR(p) = 1/N /* initial value */ while PageRank has not converged do sinkPR = 0 foreach page p in S /* calculate total sink PR */ sinkPR += PR(p) foreach page p in P newPR(p) = (1-d)/N /* teleportation */ newPR(p) += d*sinkPR/N /* spread remaining sink PR evenly */ foreach page q in M(p) /* pages pointing to p */ newPR(p) += d*PR(q)/L(q) /* add share of PageRank from in-links */ foreach page p PR(p) = newPR(p) return PR

In order to facilitate the computation of PageRank using the above pseudocode, one would ideally have access to an in-link respresentation of the web graph, i.e., for each page p, a list of the pages q that link to p.

Consider the following directed graph:

We can represent this graph as follows:

A D E F B A F C A B D D B C E B C D F F A B Dwhere the first line indicates that page A is linked

- Implement the iterative PageRank algorithm as described above.
Test your code on the six-node example using the input
representation given above. Be sure that your code handles pages
that have no in-links or out-links properly. (You may with to
test on a few such examples.) Your task will probably be easier if
you don't require loading the entire link graph into memory.
*To hand in:*List the PageRank values you obtain for each of the six vertices after 1, 10, and 100 iterations of the PageRank algorithm. - Download the in-links file for the
WT2g
collection, a 2GB crawl of a subset of the web.
This in-links file is in the format
described above, with the destination followed by a list of source
documents.
Run your iterative version of PageRank algorithm until your PageRank values "converge". To test for convergence, calculate the perplexity of the PageRank distribution, where perplexity is simply 2 raised to the (Shannon) entropy of the PageRank distribution, i.e., 2

^{H(PR)}. Perplexity is a measure of how "skewed" a distribution is --- the more "skewed" (i.e., less uniform) a distribution is, the lower its preplexity. Informally, you can think of perplexity as measuring the number of elements that have a "reasonably large" probability weight; technically, the perplexity of a distribution with entropy*h*is the number of elements*n*such that a uniform distribution over*n*elements would also have entropy*h*. (Hence, both distributions would be equally "unpredictable".)Run your iterative PageRank algorthm, outputting the perplexity of your PageRank distibution until the perplexity value no longer changes in the units position for at least

*four*iterations. (The units position is the position just to the left of the decimal point.)For debugging purposes, here are the first five perplexity values that you should obtain (roughly):

183811, 79669.9, 86267.7, 72260.4, 75132.4 *To hand in:*List the perplexity values you obtain in each round until convergence as described above. - Sort the collection of web pages by the PageRank values you obtain.
*To hand in:*List the document IDs of the top**50**pages as sorted by PageRank, together with their PageRank values. Also, list the document IDs of the top**50**pages by in-link count, together with their in-link counts. - Examine the top
**10**pages by PageRank and in-link count in the Lemur web interface to the collection by using the "e=docID" option with database "d=0", which is the index of the WT2g collection. For example, the linkhttp://fiji4.ccs.neu.edu/~zerg/lemurcgi_IRclass/lemur.cgi?d=0&e=WT04-B22-268 will bring up document WT04-B22-268, which is an article on the Comprehensive Test Ban Treaty.

*To hand in:*Explain why these documents have high PageRank values, i.e., why is it that these particular pages are linked to by (possibly) many other pages with (possibly) high PageRank values. Are all of these documents ones that users would likely want to see in response to an appropriate query? Which one are and which ones are not? For those that are not "interesting" documents, why do they have high PageRank values? How do the pages with high PageRank compare to the pages with many in-links? In short, give an analysis of the PageRank results you obtain.