Audiovisual content description using semantic scene representation

Abstract

This article describes the development of software to provide semantic descriptions of audiovisual content. Through its reliance on Semantic Web technologies, this new software will achieve a much higher level of formalisation than systems based on free text. The article describes the first deliverable of the project — the AudioVisual Semantics (AVS) ontology — which addresses the formal representation of the descriptions. The ontology has been modelled using OWL and SKOS vocabularies so that the description elements can be extended more easily. The ontology is now in the process of final validation. It will then be implemented in the VSN media asset management system to test its effectiveness and speed for cataloguing audiovisual content.