Quiz 8

Question 1

1. Which of the following is NOT a characteristic of the random walk approach to PageRank?
   A. It is a non-deterministic algorithm
   B. It is a simulation of how users navigate the web
   C. It requires a large number of iterations to converge
   D. It produces accurate PageRank scores for all web pages

Answer 1

D. It produces accurate PageRank scores for all web pages

Question 2

2. What are the benefits of using a teleport set?
   A. It can improve the accuracy of PageRank scores for certain types of webpages
   B. It can make PageRank scores more robust to changes in the web graph
   C. It can make PageRank score more computationally efficient to calculate
   D. It can make PageRank score more scalable to large web graphs

Answer 2

A. It can improve the accuracy of PageRank scores for certain types of webpages

Question 3

3. How is the teleporting approach different between topic-specific page rank and random walk with restarts?
   A. Both algorithms teleport to the initial node of the graph
   B. Random walk with restarts teleports to a single node whereas topic specific page rank cannot teleport
   C. Random walk with restarts teleports to a single node whereas topic specific page rank teleport to multiple nodes
   D. Random walk with restarts teleports to multiple nodes of the graph whereas topic specific page rank can teleport to only a single node.

Answer 3

C. Random walk with restarts teleports to a single node whereas topic specific page rank teleport to multiple nodes

Question 4

4. How does Pinterest use PageRank to rank pins in search results?
   A. It assigns a PageRank score to each pin based on the number of repins it has
   B. It considers the quality of the pins that a user has visited in the past
   C. It analyzes the content of the pins, such as keywords and descriptions, to determine their relevance to a given query
   D. It analyzes the occurrence of search queries and calculates the page rank directly

Answer 4

B. It considers the quality of the pins that a user has visited in the past

Question 5

5. True | False Topic-specific PageRank is primarily useful for search engines when ranking results for general queries.

Answer 5

True

Question 6

6. True | False PageRank is a perfect measure of web page importance and cannot be manipulated by spam farms.

Answer 6

False

Question 7

7. True | False While using spam mass estimation technique to combat link spam, picking the seed set manually is an essential step of the process.

Answer 7

True

Question 8

8. True | False In the Pixie Random Walk, the early stopping criterion is based on the number of steps taken by the random walker.

Answer 8

True

Question 9

9. How do spammers use link farms to increase the page rank of other web pages?

Answer 9

Spammers use other webpages in which they also own to promote their target website.

In which sometimes other webpages that they don’t own end up promoting it as well.

Spammer gets as many links as possible from the accessible pages to grow.

Question 10

10. Consider the following Graph
    Compute the PageRank of each node in the figure above for 2 iterations, assuming b = 0.9. Assume only nodes 4 and 5 as trusted pages. (Please not that there is a b modifier)

Answer 10

Go over it

Question 11

11. Write a pseudocode for the Pixie Random Walk algorithm and describe how similarity is computed between the nodes of a graph.

Answer 11

node = query_node.sample()
for i in range (steps)
board_node = node.get_random_neighbor()
pin_node=board_node.get_random_neighbor()
pin_node.count += 1
if random() < alpha
pin_node = query_node.sample()