Modeling Word Selection in Predictive Text Entry

  • Authors:
  • Hamed H. Sad;Franck Poirier

  • Affiliations:
  • VALORIA lab., Université de Bretagne-Sud, Vannes, France;VALORIA lab., Université de Bretagne-Sud, Vannes, France

  • Venue:
  • Proceedings of the 13th International Conference on Human-Computer Interaction. Part II: Novel Interaction Methods and Techniques
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

Although word, or short phrase, selection from a list is an extensively used task on different types of interfaces, there is no accepted model for its execution time in the literature. We present a Keystroke Level Model (KLM) adapted for this task on both desktop and handheld interfaces. The model is built from the results of an empirical study for exploring the effect of the list sorting and the number of words displayed to the user on the selection time of a word at a specified position in the list. The resulting model is integrated into an existing model for predicting the text entry speed of word-disambiguation entry methods. The predicted text entry speed is then compared to the published empirical results and to the theoretical results of the model that doesn't take into account word selection search time. The error between prediction and empirical-results is effectively reduced using the modified model.