SVG text experiment - Generative poetry restructured
This visualization presents a generative poetry interface where users construct sentences by selecting sequential textual dimensions from dropdown menus. The system loads word categories from a CSV file via `d3.csv` and dynamically creates dropdown menus using D3’s selection and data-binding methods. Each sentence selects random words from the chosen categories, with a length adjustable through an input field. The interface renders twenty generated sentence variations in HTML paragraphs, producing associative, often thought-provoking lines from the taxonomic word data.
AI-generated descriptionGenerative Text
Plan
- Have artist create a taxonomy of meaning for the words used in his texts
- Structure data to allow for easy manipulation and selection
- Create an interface that allows artist to play around with generative features
interface
- Host a simple js+html+css interface (on vizhub?)
- X-number of dropdown menus for each textual dimension
- Y-number of generated texts in simple html
- Additional features like randomization, length settings (total length but also lenght of sentences)
Changelog
- Added a sentence length input field
- Converted to a new data structure that allows easier data input by the artist in the back-end
Interesting output examples
While most of the sentences seem nonsensical, there are some thought provoking lines in the mix. Because your thought process is associative it will look for meaning and relevance and qickly find links.
see fragile phone hope loan.
why sad consumer heal superior?
sea fragile raft terror debt.
how fear country fragile tribe?
diary war memory cinema.
how religion queer ← rage?