I have three ideas this year, and this is one of them. I'm making separate issues for each.
As you may know, I've started a bit of research on NaNoGenMo, which includes a catalog (i.e. some basic metadata) of every completed entry since 2013, as well as an archived copy of the full text, if I could get it.
All told, that's 382 novels that I have the full text of, which all together contain about 45 million words.
So for this entry, I want to make a NaNoGenMo retrospective by using this corpus as the input to some of the common Travesty-esque text generation methods used by participants over the years.
To get to 50K, I think I'll do, chapters of may 8000 each, something like:
Chapter 1 Travesty
Chapter 2 Markov Chains
Chapter 3 Char-RNN
And depending how many techniques I end up with, I could do multiple chapters of some techniques and just change the parameters.
I'd also like to generate some literary criticism about NaNoGenMo, so I'm looking at SciGe and the Pomo generator to see if I can get somewhere by tweaking the nouns. I could also markovify some Harold Bloom (for example) but switch out the examples in his prose with examples from NaNoGenMo.
I have three ideas this year, and this is one of them. I'm making separate issues for each.
As you may know, I've started a bit of research on NaNoGenMo, which includes a catalog (i.e. some basic metadata) of every completed entry since 2013, as well as an archived copy of the full text, if I could get it.
All told, that's 382 novels that I have the full text of, which all together contain about 45 million words.
So for this entry, I want to make a NaNoGenMo retrospective by using this corpus as the input to some of the common Travesty-esque text generation methods used by participants over the years.
To get to 50K, I think I'll do, chapters of may 8000 each, something like:
And depending how many techniques I end up with, I could do multiple chapters of some techniques and just change the parameters.
I'd also like to generate some literary criticism about NaNoGenMo, so I'm looking at SciGe and the Pomo generator to see if I can get somewhere by tweaking the nouns. I could also markovify some Harold Bloom (for example) but switch out the examples in his prose with examples from NaNoGenMo.