This is a web app that generates new text from the works of deceased writers. There are currently 7 authors on the site:
- The Brother's Grimm
- Edgar Allan Poe
- Oscar Wilde
- Virginia Woolf
- Lewis Carroll
- William Shakespeare
- HP Lovecraft
The new text is generated using Markov models in public/scripts/tweet.js and can be tweeted from the web app to the 'dead-writers' Twitter account.
The original code was written in Python as part of a class project. The repo for the class can be found here: https://github.com/jamiejamiebobamie/tweet-generator.
This project is no longer in active development. If you have any questions or comments regarding the app, please contact me at jmccrory@vt.edu.
Click here for the live site
- Javascript ES5
- Node.js/Express.js
- Bootstrap and Handlebars
- Jamie McCrory
This project is licensed under the MIT License - see the LICENSE.md file for details
- Everytime the site launches, there is a text cleaning process that happens before the Markov model is built. Currently this process does not remove newline characters: '/n' which end up as their own tokens within the histogram. To create a better Markov model, newline characters should have been removed.
- Punctuation marks should have been tokenized. Right now punctuation is part of the words they appear next to. So for example: "girl" and "girl!" are considered to be two different tokens.
These two changes would have greatly improved the quality of the tweets and remain the focus of future efforts.
A Markov model is "a stochastic model used to model randomly changing systems" in this case the order of words an author uses.