Centrality: Generative Lexicographical Poetry

photoMeaning often lies in seemingly random events, in the chaos and messiness of life. Yet when we read, we try to get at the meaning by digging deeper, getting to the core or the center of the author’s message.

Centrality uses the Wordnet lexicographical corpus, Tristan Tzara’s dadaist cut-up technique, and Jabberwocky-esque nonsense to write a poem modeling the meaning of a single user-inputted verb. Using the very common statistical measure of centrality, the normal distribution, the poems juxtapose difference usages or senses of the verb with each other and nonsense. The effect hopefully invites the reader to treat text and language more intuitively, allowing context instead of deterministic definitions guide their understanding.

Use Centrality to generate a poem for yourself!

Read more in depth about the procedure, written in Python, here and get the code on Github.