Visualizations and Networks

network graph of selected Hum division faculty

Humanities researchers work with extremely varied data: words and phrases in text, names of individuals, images, monetary transactions, musical notation, literary characters, and others all compose the core of research material. With the use of tools to create graphs, maps, and interactive charts from this data, humanities scholars have new ways of observing and studying their research material and illuminating hidden patterns or lines of inquiry.

Creating visual and graphical representations of research material can often require varying degrees of expertise in data cleaning and programming; scholars without this background may need to seek assistance from other specialists. However, a number of tools have been developed to serve scholars without that expertise. Tableau is a commercial program for designing charts, graphs, maps, and other representations of research material and findings. Researchers aiming to get a general sense of their texts may benefit from Voyant, which allows users to load any text and perform basic analyses of word frequencies. The experimental Visual Text Explorer, developed by the Research Computing Center, generates graphical views of the placement of specified terms in a text, allowing for simultaneous close-reading and distant analysis. Those with non-textual material, such as images or geographic points, can use Palladio for similar quick-look analysis.

Some programs specialize in network analysis, or creating a chart of interlinked points—or nodes—to show relationships between data: individuals, organizations, characters, etc. Gephi can be downloaded free of charge as a basic tool for creating these network visualizations, and is frequently used. The more powerful D3 is capable of creating networks as well many different types of charts or maps, including interactive visualizations. However, it requires more familiarity with programming fundamentals and the Javascript language.

Image: 
Network graph created with D3 javascript library