The effects of word prediction on communication rate for AAC

Abstract

Individuals using an Augmentative and Alternative Communication (AAC) device communicate at less than 10% of the speed of "traditional" speech, creating a large communication gap. In this user study, we compare the communication rate of pseudo-impaired individuals using two different word prediction algorithms and a system without word prediction. Our results show that word prediction can increase AAC communication rate and that more accurate predictions significantly improve communication rate.