Personalized PageRank
In this exercise, you will study the difference between the PageRank and the personalized PageRank algorithms.
You can use the function boxplots, which shows the score distributions of churners and non-churners with two separate boxplots.
The function has two arguments:
damping, indicating the value of the damping factor. The default value is set to 0.85.personalized, a Boolean parameter that indicates whether the personalized PageRank algorithm should be used. When TRUE, the restart vector has 1 for the churners in the network and 0 for the non-churners. The default value is FALSE, i.e. not personalized.
This exercise is part of the course
Predictive Analytics using Networked Data in R
Exercise instructions
- Apply the function
boxplotsto see the distribution of the standard PageRank scores with damping factor 0.85. - Apply the function
boxplotsto see the distribution of the personalized PageRank scores with damping factor 0.85. - Apply the function
boxplotsto see the distribution of the standard PageRank scores with damping factor 0.2. - Apply the function
boxplotsto see the distribution of the personalized PageRank scores with damping factor 0.99.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Look at the distribution of standard PageRank scores
boxplots(damping = ___)
# Inspect the distribution of personalized PageRank scores
boxplots(damping = ___, personalized = ___)
# Look at the standard PageRank with damping factor 0.2
boxplots(damping = ___)
# Inspect the personalized PageRank scores with a damping factor 0.99
boxplots(___, ___)