Reflective Text Entry: A Simple Low Effort Predictive Input Method Based on Flexible Abbreviations

Abstract

Users with reduced physical functioning such as ALS patients need more time and effort to operate computers. Most of the previous assistive technologies use prefix based predictive text input algorithms. Prefix based predictive text entry is suitable for languages such as English where the average word length is approximately 5 characters. Other languages such as Norwegian and German have longer mean word lengths as words are combined into longer compound words and prefix approaches are thus less effective. This paper proposes a new abbreviation expansion algorithm. Users mentally determine an abbreviation of the word, typically comprising significant consonants and the system proposes words that contain the matched characters. The approach is non disruptive in that it does not require the user to learn a new system or abbreviation mnemonics, and it can be used with any text input device. The system is dynamic and adapts to the users style of abbreviated input. The algorithm is easier to implement than previous approaches and no a priori system training is required. Our experimental evaluations demonstrate that the algorithm achieves real time performance with modest computer hardware.