International Symposium: Programs: Summary: Session M10E

While the use of speech-to-text services is growing rapidly in educational, employment, and community settings, there is still some confusion about what meaning-for-meaning services really are. A common misconception is that meaning-for-meaning transcripts are just a summary or "dumbing down" of the message. In reality, a meaning-for-meaning transcript contains a concise and thorough message, richly detailed and in full English grammar. This presentation will discuss meaning-for-meaning services in depth, including: defining and describing the methodology and the output, exploring examples of meaning-for-meaning transcripts in relation to the verbatim text, providing strategies for service providers, and examining methods to evaluate transcripts. Participants will see examples of how voice tone and body language can be represented in a meaning-for-meaning transcript, and thus more accurately present a speaker's intended message than a verbatim transcript of the spoken English. The cognitive demands on sign language interpreters and those on meaning-for-meaning speech-to-text service providers are very similar.
The presentation will explore these similarities by using models of interpreter processing to analyze the task of producing meaning-for-meaning transcripts in real-time. Based on these analyses, the presenters will offer strategies for service providers to improve their abilities to assimilate a spoken message into a concise, clear, and grammatically correct written message. There are not only misconceptions about what meaning-for-meaning services, also are misconceptions about how the quality and accuracy of meaning-for-meaning transcripts can be measured. The presenters will discuss the rich history of objective measurement of other meaning-for-meaning areas, including foreign language interpreting, sign language interpreting, narrative writing, and instructor narrative analysis. Details of research procedures used by both C-Print and TypeWell to quantitatively measure completeness, accuracy and readability of meaning-for-meaning transcripts will be discussed. In addition, participants will be given suggestions for a streamlined method to objectively measure those same qualities in their own meaning-for-meaning transcripts.
Learning Objectives
Participants will examine the production and application of meaning-for-meaning speech-to-text services.
Participants will review methods of doing quantitative analysis of meaning-for-meaning speech-to-text services.