Dasher (software)
| Dasher
|
|
|---|---|
|
|
| Basic data
|
|
| developer | The Dasher Project |
| Current version |
4.11 ( March 14, 2010 ) |
| operating system | platform independent |
| programming language | C , C ++ , Java |
| category | Accessibility |
| License | GNU GPL ( Free Software ) |
| inference.phy.cam.ac.uk/dasher/ | |
Dasher is a graphical predictive text entry system that enables users to efficiently enter text through a pointing device using continuous character gestures rather than a keyboard . It can thus be used for the barrier-free control of computers or when the use of a usable keyboard z. B. due to the size or the intended use (e.g. for one-handed operation or operation without manual use) of the device is difficult or not possible.
The program can be used for all written languages.
It is reminiscent of an arcade game , as you drive through characters flying across the screen when you use it. It uses a probability-based prediction model to put more likely characters (combinations) in the foreground. The principle can be described as a reverse version of arithmetic coding , a data compression technique. It also uses Fitts' law , which says that larger areas can be selected more quickly than smaller ones.
Furthermore, work is being carried out on a version that combines the Dasher principle with speech recognition in order to significantly increase the efficiency of text input. For this purpose, a word grid is generated whose probabilities are based on possible ambiguities / errors on the part of the speech recognition software.
Dasher is available as free software under the GPL for Linux , Windows , Android and Apple iOS, among others . An older version is also available for Pocket PCs .
Dasher was invented by David JC MacKay and developed by David Ward and other members of MacKay's research group at Cambridge University . The Dasher project is supported by the Gatsby Charitable Foundation .
literature
- David JC MacKay: Information theory, inference and learning algorithms. CUP 2003, ISBN 0-521-64298-1 , (also available online )
Web links
Individual evidence
- ↑ a b c www.openhub.net .