Database Reference
In-Depth Information
Chapter 4
Analyzing Twitter Data
So far we have discussed the collection and management of a large set of Tweets.
It is time to put these Tweets to work to gain information about the data we have
collected. This chapter focuses on two key aspects of Twitter data for data analysis:
networks and text.
When analyzing Twitter data, we can ask many questions. Who is the most
important? What are people talking about? How are they responding to a product?
In this chapter we will discuss how to answer these questions via data analysis.
4.1
Network Measures
Many of the questions that we ask of our Twitter data can be answered through
network analysis. Questions such as “who is important?”, “who talks to whom?”,
and “what is important?” can all be answered through a network. Using proper
network measures, we can find these important actors or topics in a network.
4.1.1
What Is a Network?
A network
1
is a set of vertices linked by a set of edges. While this representation
is very simple, the choices made when creating the network can make a huge
difference in the way it is interpreted.
4.1.1.1
Vertices
Vertices are the elements that comprise a network. In both networks presented in
Fig.
4.1
, “Alice”, “Bob”, and “Carol” are the vertices in the network. Vertices can
