A little bot, coded in Python, that auto-generates biblical sounding utterances based on the King James Version Bible and then tweets them to @kjvbot via Twitter's API.
This project, which is now forking in several new directions, is a collaboration with Michael Hemenway, Justin Barber, and the NLP Working Group at Iliff School of Theology (including fellow learners Pamela Eisenbaum and Micah Saxton). It began as a much simpler program that built its utterances by mashing up formal elements and parts of speech from the text of the KJV. The latest version is something completely different, building its utterances all on its own as Markov chains based on randomly selected three-word start phrases and a selection from the KJV.
There is also a web app, built in Python and Flask, that allows users to take advantage of the same program to generate their own biblical utterances and tweet them to @kjvbot if they feel like it. Go play with the app at apps.iliff.edu/nlp/kjvbot. Have fun! And think about it!
A repository of the code and related files is available on Github, here.
Possible next steps: add part-of-speech recognition to the process (the current bot does a lot of comma splicing, which drives its professor-creators crazy); build a fuller version of the program that allows users to determine their own start words or phrases; make the Python code available for others to use with other texts or textual corpuses.