Network analysis software is another tool frequently used in the digital humanities. Visualization through networking is an effective way to present the connections and relations contained within big data to discover patterns and trends hidden in text files. Palladio is an example of network analysis tools.
Palladio is a free browser-based tool for exploring relationships developed by the Humanities + Design Lab at Stanford University. This networking software was created for scholars working in the humanities to filter and quickly produce diagrams that display data spatially and temporally through maps, graphs, tables, and galleries very familiar in this field. The visualization of text files through networking might bring new research questions waiting to be explored and thus enhance the existent scholarship.
I will walk you through the steps I followed using Palladio with the dataset from WPA Slave Narratives. This collection comprises more than two thousand interviews with former slaves from seventeen states was collected in the years 1936-1938 by the staff of the Federal Writers’ Project of the Works Progress Administration. Those interviews are available in the public domain as images and uncorrected OCR in the Library of Congress and transcriptions in the Project Gutenberg.
Load Your Data
Go to Palladio http://hdlab.stanford.edu/palladio/
First, click “Start,” which will take you to a screen where you can drag and drop the “Interviewed Palladio” file into the window & click “Load.” The first file you upload, which is your first table, becomes your primary table. Then, you can provide a name for the project by clicking on “Provide a title for this project.” The title that I used was “Slave Narratives.” To name your table, click on “Untitled” and write the name you want. In this case, the name of the table is “Interviews.”
To add another table, hover your mouse over the field that will connect to the new table, “Where Interviewed,” and the edit tool will appear. Then you click on “Add a new table” and go through the same process as before to upload “Palladio Places.” Name this table “Locations.”
Go through the same process to add the last table, “Enslaved Palladio, this time selecting “Interview Subject” to be the connector to the new table “Enslaved.” In this table, click on “Extension” and select “Location.” And, finally, click “Done.”
To adjust the mapping options, go to “Map” on the top left. This will open a new window where you will click on “Add a New Layer.”
In “Map Type,” select “Point to Point.
In “Source Places,” select “Where Interviewed”
In “Target Places,” select “Where Enslaved.”
In “Tooltip label,” select “Place Names.”
Click on “Graph” on the main menu options to create a network diagram. Here you are going to enter the parameters that are going to define your diagram. The source and the target will be the nodes, while the links connecting these nodes are the edges. You can control the way that this diagram will look. For instance, if you want the size of the nodes to reflect the number of interviews, click on “Size nodes.” To make the graph easier to read, you can highlight one of the two types of nodes, the source or the target. Take into account that this graph is interactive; therefore, you can drag and move the nodes around to make the relationships more transparent.
- The graph shown below illustrates the usefulness of networks to visualize bulks of data and how it could ignite new discussions, invite new research, and even generate new knowledge. It is extremely interesting to see how the topics of the interviews varied depending on the type of slave being interviewed.
2. The second example shows that networks do not work well to represent all sorts of relationships. There are as many interview subjects as topics, thus, the resulting graph is hard to read and the insight provided by the graph does not add anything exciting to the interviews.