Continuous edgewrite: dictionary-based disambiguation instead of explicit segmentation by the user

  • Authors:
  • Benoît Martin;Poika Isokoski;Gregory Karmann;Thomas Rollinger

  • Affiliations:
  • Université de Lorraine, Metz Cedex, France;University of Tampere, Finland;Université de Lorraine, Metz Cedex, France;Université de Lorraine, Metz Cedex, France

  • Venue:
  • Proceedings of the International Working Conference on Advanced Visual Interfaces
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

Edgewrite is a text entry method where the user follows the edges of a physical guiding rectangle to enter corner sequences that are interpreted as characters. The original Edgewrite character set resembles the Latin alphabet and includes explicit character segmentation by lifting the stylus (or centering the joystick, etc). We present a variant of Edgewrite that we call the continuous Edgewrite. It relies on a dictionary instead of user's character segmentation to disambiguate words. New users can use the continuous Edgewrite with the help of an interactive visualization of possible continuations while writing. In a 6-session user study we measured initial text transcription performance (increased from 1 to 5.4 wpm) and the ratio of observed explicit segmentations to optimal continuous writing (decreased from 2.5 to 1.5). These results show that it is possible to learn to use the continuous writing mode, but also that the learning takes some time.