I should compile a second database for my DIY Markov chain. I stripped the punctuation and capitals from the data in the first database, resulting in suboptimal generations. Using the data raw should make more interesting results.

comments