Dr Nick Hawes

Reader in Autonomous Intelligent Robotics

School of Computer Science
University of Birmingham
Edgbaston, Birmingham, B15 2TT
United Kingdom

Email: n.a.hawes@cs.bham.ac.uk
Twitter: @hawesie
Phone: +44 (0) 121 41 43739
Office: 133 (first floor, back right)
Office Hours: Mon 12:00, Tues 11:00 (term-time only)
Availability: Doodle MeetMe
[hawes04cst9] Nick Hawes and John Kelleher. Context-Sensitive Word Selection For Single-Tap Text Entry. In Eva Onaindia and Steffen Staab (editors) Proceedings of the Second Starting AI Researchers' Symposium (STAIRS 2004), volume 109 of Frontiers in Artificial Intelligence and Applications, pages 217--222, IOS Press. August 2004. [pdf] [bib]
Abstract.

Predictive text input using a single-tap entry method is currently the standard for text entry in the mobile domain. One problem facing this approach is which word to present to the user when more than one word matches an input sequence. The standard single-tap approach selects words using corpus occurrence frequencies, ignoring linguistic context. This can lead to the selection of words that are unrelated to the current input. In this paper we present an implementation of a context-aware predictive text framework that overrides the frequency-based selection model for words related to the current context.

Download:  pdf pdf (84 KB)  bib bib
Links: [Google Scholar] [CiteSeer]