You are on page 1of 2

1. Draw the ego network for Node 4 using shortest path of length 2.

How
many alters are in this network?

2. What is the degree centrality of Node 4?

(NB: to make it easier for this question only, the unnormalised version is fine,
i.e., no need to divide by N-1)

3. What is the betweenness centrality of Node 4?

4. What is the closeness centrality of Node 4?

5. Suppose the nodes correspond to people and the edges correspond to


friend relationships. As an advertiser you have knowledge of this network
and you have the budget to target a single person with an advertising
campaign. Your hope is that if you target the right person, then knowledge
of your product will spread by word of mouth via the friend relationships.
Provide three reasons why node 4 is the best person to target.

6. Suppose the nodes correspond to people and the edges correspond to


friend relationships for a social network. As a social network provider you
decide to make a recommendation to the person corresponding to node 2.
You want to recommend an additional node for it to connect to (via an
undirected edge), in order to achieve a maximal increase in its betweenness
centrality. Which node should you recommend and why?

7. What is the largest 2-clique in the network in Figure 1 (ie what is its size
and which nodes participate in it)?

8. Write the adjacency matrix for the network in Figure 1.


9. What is the modularity score of the following partition of the network into
two communities: {1,2,3,4}, {5,6,7,8,9}? What is the modularity score of the
network partitioned as a single community: {1,2,3,4,5,6,7,8,9}? Based on
your answers, explain which of the two partitions is a preferable
community structure for this network and why.

10. Consider the following partition of the network in Figure 1 into 9


communities: {1},{2},{3},{4},{5},{6},{7},{8},{9}. Explain intuitively, why this
partition is unlikely to serve as a good community structure.

You might also like