Palladio is a network visualization tool developed by Stanford University’s Humanities + Design Lab, which allows researchers to “visualize complex historical data with ease.” It’s designed for humanities research, which includes complex data such as age, gender, people, places, and many other “nodes”. It is used to create visualizations of patterns and relationships among all these different node types.
Palladio is browser-based. Tabular data is uploaded; tutorials and FAQs on the Palladio website provide clear directions about the kind of data Palladio accepts and how it should be structured.
The first step is to figure out what data you want to include in your table, and to input that data directly into Palladio, or create a .txt or .csv file that can be dragged into the interface. It’s important to read the guide about how the data should be input into the cells; for example, if you have multiple values (as in one of the tables for our Module 9 activity, “house, field”) in one cell, you have to use the control in the window for that table to let it know that the comma is a “delimiter string”, and it will separate those two values.
A main table is loaded, followed by other tables. The main table is the base, it’s the one that has the most categories or variables. Other tables can be extended, or linked, to it. In our example, we linked “Interview Subject” to a table called “Enslaved”. It will then match the interview subject in table 1 to the enslaved person in table 2. Once you have loaded all your tables, you are now ready to create visualizations–maps, graphs or tables. Each type of visualization has a settings panel for further control. Maps allows for cartographic visualizations, where the nodes are shown as points, or point to point (showing a flow between the two, as a letter, or a voyage, etc.) The one of most interest to scholars is Graph. One can visualize patterns and relationships by correlating a source and a target. For example, if you choose “interviewee” as a source, and “where interviewed” as the target, the visualization will show the node for the location, with nodes for each person radiating and connected to the location by a line. If you select “size nodes”, the location nodes will be different sizes, according to the number of interviews. One can filter and focus the data further by means of the Facet control at the bottom. When learning, it’s best to experiment with uploading tables and then playing with the source vs. target controls as well as the extensions to get a sense of the patterns and how this information is visualized.
Palladio is easier for scholars to learn and understand than more complex software like Gephi. The controls are simple enough while still allowing for complex patterns and relationships based on all kinds of combinations of data.
Because it is web-based, the project must be downloaded to be saved. Visualizations can be exported, but only as static images rather than interactive visualizations, as in the case of Voyant.
Below are examples of exported visualizations, using data from the WPA Slave Narratives.
Palladio graph visualization resulting from interviewer as source, interviewee as target
Palladio point to point map visualization linking where interviewed to where enslaved
