Spoken Language Interaction in a Goal-Directed Task

<p>To study the spoken language interface in the context of  a complex <br>problem-solving task,  a  group  of  users  were asked  to  perform  a <br>spreadsheet task, alternating voice and keyboard input.  A total of 40 <br>tasks were performed by  each participant, the first thirty in  a group <br>(over several days),  the  remaining ones a month later.  The voice <br>spreadsheet program used in this study was extensively instrumented <br>to provide detailed information about the components of  the inter- <br>action. These data,  as well  as analysis of  the  participants's  ut- <br>terances and  recognizer  output, provide a fairly detailed picture of <br>spoken language interaction.</p> <p><br>Although task  completion  by  voice took  longer than  by  keyboard, <br>analysis shows that users would be  able to perform the spreadsheet <br>task faster  by  voice,  if  two key  criteria could be  met  recognition <br>occurs in real-time, and the error rate is sufficiently low.  This initial <br>experience with a spoken language system also allows us to identify <br>several metrics,  beyond  those traditionally associated  with  speech <br>recognition, that can be used to characterize system performance.</p>