Design rationale and evaluation of the Portable Anticipatory Communication Aid—PACA

Publisher Website
Google Scholar
Abstract
The Portable Anticipatory Communication Aid (PACA) was developed as a prototype predictive communication aid to evaluate the efficacy of anticipatory techniques in actual communication situations by augmentative and alternative communication (AAC) system users. The anticipatory techniques, which include both letter and word anticipation, are directed at scanning methods of message production. The objective of these techniques is to reduce the number of equivalent switch activations needed to produce a message. Letter anticipation is based on the probabilities of letter pairs (digrams). The letters in the selection array are organized by their likelihood of following the last selected letter (or space). Letters of higher probability are placed in positions within the array which require fewer equivalent switch activations to select. Letters of lower probability are placed in those positions which require more equivalent switch activations. Each time a letter is selected, the array is reorganized to reflect the changed digram probabilities for the selected letter. Once the user selects the initial two letters of a word, the PACA presents a group of words as predictions for completing the word. In the prototype PACA the words were derived from a fixed set of core words and a user-defined set of learned words. The core words represent words which have generally high frequencies of occurrence in American English. The learned words include words particular to the individual user and to the topics and situations of that user. Learned words are automatically incorporated into the word list when the user first spells them in the course of creating a message and are automatically discarded if not utilized. In the selection groups, the words are ranked by their frequency of use. The PACA was implemented on a commercial portable computer, the Epson HX-20, and provided with features appropriate for augmenting conversational communication, note-taking, and writing. Three individuals, all with severe physical impairments and users of augmentative communication systems, were selected to evaluate the implementation of the anticipatory techniques in their dayto- day communication needs. Personal observations of the evaluators were complemented with formal observations of their operation of the PACA and analysis of changes in their learned vocabularies. The impressions of the evaluators supported the general efficacy of anticipatory techniques for improving the efficiency of message production. Analysis of word use and the makeup of the learned vocabularies led to several changes in the presentation, manipulation, and structure of the vocabulary used for word anticipation.

This publication has 1 reference indexed in Scilit: