I have to say that text mining was rather less fun than I had expected. Yes, the Ngram and Bookworm and Voyant were fun for searching terms (see previous post), but I am having difficulty figuring out what I would use the topic modeling for. Using an (admittedly) small sample of texts, it didn’t tell me anything I didn’t already know about my research. But that is perhaps because I am fairly far advanced on this project; I’ve done the research, am familiar with all the sources, and have done some of the writing. So maybe the best time for this distant reading and topic modeling is when you are beginning a new project and want to run a bunch of texts through the old topic modeler. I am trying to think how it might’ve helped my research if I had had the ability to do this for the gift project. If I had been able to load hundreds of articles on gifts, etiquette books, diaries, etc. and see what relationships popped up in topic modeling, it could have set up some questions for research. This might have structured my research earlier and saved me some time in slogging through all these documents to come up with the key concepts to frame my argument. As it is, I think my current research is too far gone to benefit substantially from this. But perhaps there are dimensions to this that I’m not considering; I’m willing to be proven wrong on this.
For my DDH project, which is my course on DH, I would like to create an assignment using NGram and/or Bookworm, so that students would be able to see the rise and fall of words/phrases over time. I’m not quite sure what the assignment will specifically consist of at this point, but I will work on it. I think Voyant is a bit tricky, and the topic modeling even moreso, for an introductory undergrad course in DH.
I think the middle portion of my course will be “playing with digital tools.” I am hoping the students will all have laptops so that we can do these sessions in class like we have done in this seminar. If not, I will have to try to get the library classroom or one of the labs (which are in short supply, especially for humanities–one of the many problems at my campus and, I am sure, other campuses).
Source: Uses of distant reading?