Reasonable goals for a first DH project? « Digital Humanities Questions & Answers

Digital Humanities Questions & Answers » Applications, Tools, Formats

Reasonable goals for a first DH project?

(3 posts) (2 voices)

Asked 1 year ago by gscottrobinson
Latest answer from gscottrobinson

Tags:

gscottrobinson
Member
Hi!

I'm interested in learning more about computer-based textual analysis, so I developed a project using the journal Literature/Film Quarterly (LFQ). I haven't done this sort of research before, so I'm trying to determine the best questions to ask and the tools that will help me answer them. I'm hoping someone might be able to tell me if these are reasonable questions to ask. I'd also appreciate any suggestions.

LFQ began in 1973 and continues now. As the oldest journal in the field, looking at the text should be able to provide some insights about how adaptation studies developed. At about 50 articles a year for 43 years, I'm looking at 2150 articles.

Questions:

Can an examination of the entire text of LFQ reveal anything substantial about the way adaptation studies has changed from 1973-present?
What are the most frequently cited texts in each year? WebofScience doesn't support this journal.
What names (other scholars) appear most frequently each year?
What kinds of terms are used in earlier years (e.g. “fidelity,” "intertextuality") and do they continue throughout the journal’s history?

If there are other questions that I could explore, I'd appreciate any suggestions.
Tweet this question
Posted 1 year ago Permalink
Patrick Murray-John

Goals and scope for the project might depend on the difficulty of the part of the project that's implicit in all your questions -- getting OCR'd text or some other electronic form out of the articles. A quick look at their site makes it look expensive, and some publishers get squeemish, at least, about this kind of project.

Depending on how you can get reasonably in machine-readable form, you might have to limit the scope to only one issue a year, or to a smaller date range, or both. That might make the 4th question a bit more feasible than the others.

On the other hand, if you can get everything, the 1st question sounds like a good topic modeling project.

Getting cited texts sounds pretty hard, or at least labor intensive. It would be difficult to parse out citations from the plain text. However, if you can't get plain text and only have images to work from, then that sounds like the most feasible question to pursue after you do that labor intensive work of collecting the data.

Posted 1 year ago Permalink
gscottrobinson
Member

Thank you so much, Patrick. The advice is appreciated. I've managed to download several years of the text, but the OCR is pretty terrible. Also, I've read a bit about topic modeling, but I'll look into it a bit more.

Gregory

Posted 1 year ago Permalink

RSS feed for this topic

Reply

You must log in to post.