For information about this project, please see the page on my website.

The output is available here.

The source data is available here.

The notebooks run in order:

1_select_cliched_sentences.ipynb
2_create_new_sentences.ipynb
3_make_stories.ipynb
4_get_names.ipynb
5_clean_stories.ipynb
6_make_book.ipynb

The notebooks require Mallet, gensim, Stanford CoreNLP, spaCy, and Latex. Mallet and CoreNLP require Java.

I use an older (1.9) version of spaCy (the current version seems to require a huge amount of memory for novel-length texts); I could, of course, have used CoreNLP for everything . . .