Efficiency of Limited Vocabulary Speech Recognition for Data Entry Tasks

Publisher Website
Google Scholar
Abstract
Performance using four different levels of speech entry was compared to manual entry using a keyboard. Twenty subjects entered text as individual words, phrases, and characters using both orthographic and phonetic alphabets. Overall, when text is entered as words or phrases, performance is equivalent to using a keyboard for individuals who are able to type reasonably well. Spelling text character by character produces approximately seven times as many errors and requires significantly more time to enter information. Use of a phonetic alphabet improves performance but is not as efficient as phrase and word entry.

This publication has 0 references indexed in Scilit: