A predictive text web based artwork by Simon Biggs
In these web based artworks the reader/writer can type into a text field displayed on the screen. When they/you press the space bar the software assumes they/you have entered a complete new word and searches for every occurrence of that word in a reference text (in the case of examples 1 and 2 this is the short story "The Great Wall of China" by Franz Kafka). The software then determines all the words that follow the typed word and counts how many times each occurs. This statistical data is used to predict all the possible words that might follow what the reader/writer has typed. In Predictor 1 this data is employed to write a parallel text alongside that which the reader/writer is typing. In Predictor 2 the data is used to determine the tone and size of words that are displayed near to and around the word the reader/writer has just typed.
These are prototype works for developing an installation, in collaboration with Andrew Roberts and Martin Fischer (University of Dundee) and Mark Shovman and Kenneth Scott-Brown (University of Abertay Dundee). This work combines full 3D immersive visualisation systems, speech recognition and automatic language acquisition, allowing the creation of an immersive 3D textual environment that interacts with and forms, over time, from the utterances made within it, forming a heuristic dialogue of probabilities.
A Little Pig Production
If you do not have Shockwave you might like to download it.