In reference to the reading on Topic Modeling, I’m interested in how this method might be useful in regards to pulling data from fashion publications (Vogue, Harper’s Bazaar, W etc.) Topic Modeling is defined as a form of text mining, a way of identifying patterns in a corpus. You take your corpus which groups words across the corpus into “topics.” We’ll say that the reoccurring words or “topics” are trends. If I’m applying this to my fashion research, then trends could be searched literally across 5 years worth of words collected in a magazine. What are the “trends” (words and actual style trends) that occur the most frequently from 1960 to 1965? 1975 to 1980? Were there specific topics being talked about in these chunks of time that happened to reoccur? In these chunks of time did the topics vary by publication? Did the language stay the same? I am interested in pulling data from publications, but I’m not sure if I want to pull text data or photo data (as in, maybe, how many times in a decades worth of issues from Vogue were black models featured in editorials).
Great thoughts, Scarlett. You might be interested in the Vogue project at Yale — http://dh.library.yale.edu/projects/vogue/ (news story here – http://yaledailynews.com/blog/2014/10/30/digital-humanities-go-vogue/ ) that is doing some of the things you suggest!
Yeah, I actually learned about this last year when our Culture of Fashion class went to the NYPL to learn more about the digital (and print) resources available to us. We have a great relationship with their Fashion librarians, so I hope to use them on my DH journey.