Category Archives: Relationship Mapping

Three Types of Visualization Software: Voyant, kepler.gl, and Palladio

Three open-access tools for the Digital Humanities offer different ways to analyze large collections of information.  For this class, we used the WPA Slave Narratives as digitized by the Library of Congress as source material to explore the functionality of each piece of software and its usefulness as an analytical tool.

Voyant is designed for text mining to create visual representations (graphs and word clouds) of the common terms found across a large database of source material.  It is most useful to analyze the common language and topics that occur across a large collection of source material.  It is relatively easy to use and easy to “toggle” between views to understand the prevalence of common terms across a large dataset and within identified subsets or collections within a larger dataset.  Like all of three software programs, the quality of visual representations relies on the extent of the source material and understanding how it was created.  I think this is an excellent tool for evaluating literary works and other “fixed” sources where the author’s intent is clearest.

kepler.gl is a mapping software.  I found it the most cumbersome to use as a non-techie person, although being able to generate geographic visualizations of the sources of large datasets like the Slave Narratives is very useful.  Because my interest is less in mapping patterns across a geographical area and more in the relationships of the material itself, I found this produced the least useful visualizations at least with the dataset that we were using.

Palladio is a visualization tool that focuses on relationships between datasets, such as person and location or type of worker and topics discussed.  I found this to be an easy-to-use software that was most illuminating when focusing on interpersonal and subject relationships.  For geographic relationships, kepler.gl would be more useful.  Again, because my interest tended to be more topic-oriented than location-oriented, I found the visualizations to be easy to read and to understand and useful for my understanding of the source material.

Palladio

Palladio is an easy-to-use online software that allows you to map relationships between two focus areas in a data set.  It is most useful for illuminating relationships between subjects (such as influence networks: teacher-student; mentor-mentee; etc.).  It allows a good deal of flexibility in the creation of graphs and their manipulation so the user can see quickly how these relationships played out within the data set being studied.

I found it most illuminating for the type of interpersonal relationships that Humanities scholars often wish to study–how one writer might have influenced another; musical influence networks; etc.  While geographic location could be mapped, this was mostly useful when the number of physical locations was smaller; a rendering of this information on a mapping software would be more revealing.

For my own use, being able to understand how different creators interacted with each other is very valuable–possible unknown relationships can be revealed and verified through further study of the source material.  To be truly meaningful, however, the source material should be as rich as possible; otherwise some key relationships might be missed.  Understanding the limitation of your data set is always important for making the best use of a tool like Palladio.