CAtN » Sketch 5: NaNoGenMo

1 minute read

There’s a line in something I wrote years ago:

‘The tale of a boy raised among wolves,’ someone has suggested.

In the story the boy is raised by exclusively by women. As an adult he comes to realize he’s something of an outlander.

The line itself is by no means profound but I like the premise of the thing and while perusing titles sorted by popularity on Project Gutenberg, I came across Rudyard Kipling’s The Jungle Book, the story of a boy raised by wolves. So that’s the text I choose to work with.

Python is new to me and I struggled with it, as I continue to struggle with Javascript, but I think I have a notion of what make Python so attractive. It seems like a real workhorse and I understand why it’s the tool of choice for NaNoGenMo. Jupyter Notebooks is pretty nifty, too.

I enjoyed our first assignment using Tracery and unlike the other more computationally sophisticated tools I’m able to wrap my head around it conceptually as a kind of skeletal structure on/in which to hang/infuse raw material extracted from donor texts. Working backwards from Tracery I opted to play with spaCy. Other than having spent perhaps half an hour playing with Runway ML, I’ve done absolutely nothing with machine learning.

Honestly, I had no notion of what I’d hoped to accomplish and I’m sure it shows. The problem, at least as I experience it, is one requires a certain proficiency with the tools before one can begin to plot a course of inquiry. You can see it in my code, the futile compulsion to step through spaCy’s parts of speech tagging, absent a purpose (let alone an outcome) in mind.

Code and NaNoGenMo are here.