LetterWise: prefix-based disambiguation for mobile text input

  • Authors:
  • I. Scott MacKenzie;Hedy Kober;Derek Smith;Terry Jones;Eugene Skepner

  • Affiliations:
  • York University, Toronto, ON, Canada;Columbia University, New York, NY;Eatoni Ergonomics, Inc., New York, NY;Eatoni Ergonomics, Inc., New York, NY;Eatoni Ergonomics, Inc., New York, NY

  • Venue:
  • Proceedings of the 14th annual ACM symposium on User interface software and technology
  • Year:
  • 2001

Quantified Score

Hi-index 0.00

Visualization

Abstract

A new technique to enter text using a mobile phone keypad is described. For text input, the traditional touchtone phone keypad is ambiguous because each key encodes three or four letters. Instead of using a stored dictionary to guess the intended word, our technique uses probabilities of letter sequences --- "prefixes" --- to guess the intended letter. Compared to dictionary-based methods, this technique, called LetterWise, takes significantly less memory and allows entry of non-dictionary words without switching to a special input mode. We conducted a longitudinal study to compare LetterWise to Multitap, the conventional text entry method for mobile phones. The experiment included 20 participants (10 LetterWise, 10 Multitap), and each entered phrases of text for 20 sessions of about 30 minutes each. Error rates were similar between the techniques; however, by the end of the experiment the mean entry speed was 36% faster with LetterWise than with Multitap.