The Character Thinks Ahead: creative writing with deep learning nets and its stylistic assessment

TL;DR

This paper explores controlling deep learning text models to generate stylistically distinctive poetic works, assessing the success of these controls in creating stylistic uniqueness, exemplified by their piece 'The Character Thinks Ahead'.

Contribution

It introduces methods for controlling deep learning text generation to produce stylistically distinct poetry and evaluates their effectiveness.

Findings

Controls can influence stylistic features of generated text

The approach successfully creates stylistically distinctive poetic outputs

Potential applications extend to various creative writing contexts

Abstract

We discuss how to control outputs from deep learning models of text corpora so as to create contemporary poetic works. We assess whether these controls are successful in the immediate sense of creating stylo- metric distinctiveness. The specific context is our piece The Character Thinks Ahead (2016/17); the potential applications are broad.