Automatic Annotation of XHTML Pages With Audio Components

In this paper the authors present Deiush, a multimodal system for browsing hypertext Web documents. The Deiush system is based on the novel approach to automatically annotate hypertext Web documents (i.e. XHTML pages) with browsable audio components. It combines two key technologies: Middleware automatic separation of Web documents through structural and semantic analysis which is annotated with audio components, transforming them into XHTML+VoiceXML format to represent multimodal dialog; and Opera Browser, an already standardized browser which they adopt as an interface of the XHTML+VoiceXML output of annotating. This paper describes the annotation technology of Deiush and presents an initial system evaluation.