InnsOS is a small project I started to try out some complex network analysis for humanities applications, namely character networks in literary texts such as plays. I’ve been inspired by the work done by DLiNA and thought I might as well do something similar with a rather different corpus of dramatic texts: German medieval religious plays. My favorite play is the Innsbruck or Neustift Easter Play from 1391 which is thought to have originated in my hometown of Schmalkalden/Thuringia (in the Thuringian Forest in the middle of Germany) and is quite famous for its extended non-religious, burlesque scene: The so-called Peddler’s Play.
Unlike the hundreds of neatly XML-TEI encoded and curated German and Russian dramatic texts in the Drama Corpus dracor, medieval religious plays are not available in digital format, sometimes not even in a printed edition. Thus, I have to gather the necessary data for my analyses manually - usually in form of edges lists and nodes lists that I create as csv files. They can be found in data but originate from a Google spreadsheet with significantly more information (that is not needed for CNA).
I play around with network analysis tools, like Gephi, and try my hand at CNA in Python with the NetworkX library.
You find a Jupyter notebook in which I try to follow along the book Complex Network Analysis by Dmitri Zinoviev and another notebook where I try to apply what I learned from the book to the Innsbruck Easter Play data. I’m not very good, though…
In the long run, I want to create a data model for character interaction networks for medieval religious plays, so that relevant data can be collected (manually, I suppose) in a standardized way, gathered in a centrally hosted repository and used for mid-distance ‘reading’ (i.e. analysis) of medieval plays.
Additionally, I want to use this project to teach myself CNA and especially literary network analysis so I can add this to my DH repertoire and teach others in how to do it.
Ultimately, I want to create a corpus of all Norwegian (bokmål/nynorsk) dramatic texts in the digital collection of the National Library of Norway and extract character network data in a systematic fashion. This should include the complete edition of plays by Henrik Ibsen and plays by Ludvig Holberg, too. To integrate these into larger, international drama corpora would be the final stage.
In the meantime, I will work on this project in my after work hours and publish preliminary outcomes on my weblog Digital Textology.
Written with StackEdit.