The purpose of this proof of concept application is to create computer generated haiku that is technically correct and sensible. To achieve this goal I employed the use of the Natural Language Toolkit, the CMU Pronouncing Dictionary, the Corpus Of Contemporary American English, and Python as glue for the application. Statistical models were used when tagging parts of speech and generating sequences of words. The application does not intend to approach the subjective nature of artistic expression. This is a much more complicated problem and does not fall within this program's scope. Instead, I focused primarily on the creation and editing of grammatically and syntactically correct haiku.
Advisor: Marietta Cameron