This is a technology or research stand, located in Zone R6 - Content and Knowledge. Stand number: R6-12
The systems in the CLASSIC – computational learning in adaptive systems for spoken conversation project – are learning from experience.
This exhibit presents the CLASSiC project which will display interactive speech interfaces showing the latest advances in statistical methods applied to human language technologies. This includes the ‘1013+’ industrial system for customer-service calls or automatic appointment scheduling from France Telecom/Orange labs, which is currently being deployed in France. On the stand there will also be tourist information versions of CLASSiC systems. The novelty of this project is that the technology handles uncertainty about speech interaction: “what did I hear?”, “what did it mean?”, and “what should I say next?” This uncertainty is represented by changing probability distributions which visitors will see when they try out the systems.
Coordinator: Oliver LEMON (Heriot Watt University, Mathematics and Computer Science (MACS), United Kingdom of Great Britain and Northern Ireland)