With the help of MALLET and Gephi, I walked through the process of creating topic networks and dealt a bit with their possible uses. For my corpus, I focused on papers, chapters, blog posts and collaborative works that grapple with the definition of the Digital Humanities.
I’ve also posted all the data I used if one wants to load up the resulting graphs or dig into the raw MALLET output, along with a few visualizations. The purpose of this exercise was to examine and illustrate the process and output of topic modeling, which is new to me, and so I’m sure those with more experience with topic modeling and text analysis will think my setup rather unsophisticated. Though I deal a little bit with network analytical concepts like centrality, I really am not making any claims about particular papers or blog posts to be central to the Digital Humanities, and I studiously avoided that form of analysis, which I think would be unwise given the rough nature of the thing. Still, I found the results very interesting and would love to hear any comments they might elicit.