Skip to content

Latest commit

 

History

History
7 lines (4 loc) · 1006 Bytes

README.md

File metadata and controls

7 lines (4 loc) · 1006 Bytes

Evolving-Titles

pip install -r requirements.txt

Evolving titles is a model designed to generate book titles. The components of the model represent fundamentals of human creative thinking, such as free association propagating through a semantic network of concepts and evaluation of ideas based on former experience. The model uses WordNet, an open source semantic graph of English words to generate a whole population of new title ideas, which are then improved by evolutionary algorithms. Fitness of the individual ideas in a population is evaluated by a neural network, previously trained to differentiate real book titles from rudimentary machine-generated pieces of text. Results show a significant improvement in the ability of the model to create sensible and even original titles.

The original idea of combining a neural network with evolutionary algorithms to solve a creative task is adopted from Robert Levy.