Speech Content

Our framework for developing speech-enabled CALL content extends CALL-SLT Lite, developed under a previous SNF-funded project. CALL-SLT Lite supports the construction of interactive prompt-response applications, and was among other things employed to build the application used to collect data for the Spoken CALL Shared Task.

During the current project, we are focusing on making CALL-SLT Lite applications scalable and easier to construct. CALL apps are defined using a simple spreadsheet-based format and deployable both on normal webservers and on the Amazon Alexa platform. Our initial experiments have used five sample apps, which we have run in the Alexa version; we will soon be making these generally available through the Alexa store. The table below gives details. Each app exists in three versions, for English, French, and German.