Week 1

Question 1

What are some of the challenges in building a web search engine?

Answer 1

A large number of web pages
Web pages are not ordered or indexed
No classification of web pages

Question 2

What are the two principles of the PageRank algorithm?

Answer 2

Hyperlink tricks
Random surfer model

Question 3

What is the hyperlink trick?

Answer 3

Pages can be ranked by the number of links from which this page is linked to

Question 4

What is the matrix form of pagerank?

Answer 4

P_n+1 = H x P_n

Where H is the transition matrix

Question 5

What are the two issues in PageRank?

Answer 5

Sink Pages
Cycle Pages

Question 6

How do we solve the sink page problem?

Answer 6

Replace the column of all 0s with 1/N in the P matrix

Question 7

What is the cycle problem?

Answer 7

Few webpages may be linked to a closed cycle, leading to an infinite increase of the authority score

Question 8

What is the random surfer model?

Answer 8

A surfer will randomly click on a link (accessing the site directly) or click a random hyperlink on the website.

Question 9

What does the damping factor do?

Answer 9

It’s a chance that a user will stop clicking links and get bored.

Question 10

What is the complexity of PageRank?

Answer 10

O(MN^2)
M = number of iterations