Digital Scholarship Research Guide

What is network analysis?

Network analysis is the exploration and visual representation of the connections among entities in a dataset. Often in the humanities and social sciences, network visualizations represent social networks exploring the connections among people; however, networks can also represent relationships among places, things, concepts, etc. Typically networks must consist of entities or “nodes” (often represented by dots or circles) and then the connections or “edges” among those entities (often represented by lines drawn to connect nodes).

Arrows can also be added at the end of lines between nodes to indicate the direction of the interaction, if applicable. For example, an arrow pointing from person X to person Y could indicate that X wrote a letter to Y. The edges with arrows are referred to as “directed” while edges with no arrows are “undirected.”

Other attributes of your network such as the gender, age, or occupation of a character; the genre, category, or date of a publication, etc. can be represented visually by different colors or shapes of the nodes. These visual variations can then be used to explore differences in relationships or communication frequency by one of the defined attributes.

Creating a dataset

If you are conducting network analysis on data you have identified yourself, your data needs to be formatted in a particular way in order to work with visualization software. Usually, for the most basic networks, you will need to create a list in a spreadsheet of all entities you want to include, called a “node list.” You will also need an “edge list” containing the interactions between nodes with a node name in the first column, and the node name of the other entity in the second column with whom an interaction takes place.

Finding prepared datasets

Several websites compile data that can be used to create networks graphs. Some high quality ones can be found below.