Can Automatic Abbreviation Expansion Improve the Text Entry Rates of Norwegian Text with Compound Words?

Abstract

Text entry can be challenging for users with reduced motor function. Prefix-based prediction has been successfully applied for text in English and other languages. However, prefix-based text prediction does not work optimally in Germanic languages such as Norwegian where long words are constructed through linguistic compounding. This study explored automatic abbreviation expansion as a means for improving text entry performance. A mixed text entry experiment involving n = 36 participants without reduced motor function was conducted. Three types of input devices designed for individuals with reduced motor function were used in the experiments, namely a virtual keyboard operated by a touchpad, a software head mouse, and keyboard scanning. Physical QWERTY keyboards were also included for reference. The results show that the abbreviation only lead to a marginal improvement with the slowest of the four input techniques, namely keyboard scanning. Abbreviations yield slower text entry performance with the other text entry modes.