Database Reference
In-Depth Information
4.1.2
Networks from Twitter Data
In the previous section we discussed the constructs that make up a network. We have
kept our definition of a network general to allow for flexibility in the interpretation
of the calculations performed on these networks when building them with different
definitions in mind. When we build a network, our definition of nodes and edges
determines the meaning of the measures performed. We begin by constructing a
retweet network.
The retweet network is special because it can be derived directly through the
Streaming API, there is no need to wait on the highly-limited REST APIs to
extract this network. Additionally, it carries a very intuitive meaning. The nodes are
individual users. The
directed
edges indicate the flow of information in the network.
An edge from node A to node B indicates that A has retweeted B, forwarding
B's information to his followers. Studying the structure of this network yields
information about the ways the users communicate, and how highly they value each
others' information.
4.1.3
Centrality: Who Is Important?
Often, we want to know who is the most important person in the network. The
question of importance may not be as straightforward as it seems. There are several
dimensions along which one may be considered important. Measures of importance
in social networks are called “centrality measures”. Here, we discuss the three that
are used most frequently in social media analysis. Each provides a different view of
who is important in the network. We will use the retweet network to demonstrate
these centrality concepts.
4.1.3.1
Degree Centrality: Who Gets the Most Retweets?
One of the most commonly-used centrality measures is degree centrality. The
calculation is simple: count the number of links attached to the node, this is their
degree centrality. In the case of a directed graph, there are two interpretations:
In-
Degree Centrality
, the number of edges entering the node, and
Out-Degree
,the
number of edges leaving the node. In the context of our retweet networks,
In-Degree
Centrality
is the number of users that retweet the node, and
Out-Degree Centrality
is
the number of users the node retweets.
2
In Listing
4.1
, we present an example of
In-
Degree Centrality
.InFig.
4.2
, we see an example retweet graph with the nodes sized
by their
In-Degree Centrality
. Interpreting this picture is simple, since Alice has the
most edges pointing towards her, she is the most important node in the network.
2
Notice that we have omitted an edge weight based upon number of times a retweet occurs. We
omit this for simplicity.
