You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I've played around a lot with Emmett Williams' IBM Poem generation for class projects, so I'm using it for my first NaNoGenMo project too.
for those unfamiliar the method involves assigning a word to each letter of the alphabet. Then you choose three letters and expand those into the words you assigned them, which serves as the title. Then for each word in the title, you expand their letters into a line of poetry, then expand those words in those lines, and so on and so forth.
You typically stop three loops in, but I'm throwing out the limit and writing 50k words out of a list of 26.
I was also interested in looking for ways to 'cheese' the generation to be more intelligent, specifically by combining part of speech with letter frequency. For example, E, S, D, and T are the most common letters at the end of a word, so they shouldn't be a prepositions.
There are many areas for improvement in that regard, but simply getting the first part done with is good for me.
I coded this in p5js and used Bindery to put it in a book layout, you can see the final state of the code in the repository or some possible post-2022 edits on p5js.org
The text was updated successfully, but these errors were encountered:
I've played around a lot with Emmett Williams' IBM Poem generation for class projects, so I'm using it for my first NaNoGenMo project too.
for those unfamiliar the method involves assigning a word to each letter of the alphabet. Then you choose three letters and expand those into the words you assigned them, which serves as the title. Then for each word in the title, you expand their letters into a line of poetry, then expand those words in those lines, and so on and so forth.
You typically stop three loops in, but I'm throwing out the limit and writing 50k words out of a list of 26.
I was also interested in looking for ways to 'cheese' the generation to be more intelligent, specifically by combining part of speech with letter frequency. For example, E, S, D, and T are the most common letters at the end of a word, so they shouldn't be a prepositions.
There are many areas for improvement in that regard, but simply getting the first part done with is good for me.
Here's my final result!
I coded this in p5js and used Bindery to put it in a book layout, you can see the final state of the code in the repository or some possible post-2022 edits on p5js.org
The text was updated successfully, but these errors were encountered: