INVITED TALK
Extensible MultiModal Annotation markup language (EMMA)
Max Froumentin, W3C
Abstract
This talk will introduce the W3C Multimodal Interaction Activity, whose goal is to design a
framework of specifications to enable access to the Web using multi-modal interaction. In particular we
will introduce the Extensible MultiModal Annotation (EMMA) language specification. EMMA is an
XML language for describing the interpretation of user input, combining transcriptions of raw signals
into words with metadata to help applications resolve uncertainties and contradictions in interpretations.
We will also discuss the issues encountered by the Working Group as the language is being developed,
such as whether to use RDF or not, how best to combine interpretations, or what metadata properties
to represent.

