LinkedMDR: Linked Multimedia Document Representation

A global overview

We propose a novel ontology for Linked Multimedia Document Representation, entitled LinkedMDR. This ontology models the knowledge embedded in a corpus of heterogeneous documents. It is based on (i) the combination of the relevant standards for metadata representation and content description
(e.g., DC, TEI  and MPEG-7), and (ii) the introduction of new semantic concepts and relations that augment the capabilities of these standards and link them through a semantic network.

Our proposed ontology is made of three main layers: (i) the core layer serving as a mediator, (ii) the standardized metadata layer comprising elements from existing standards, and (iii) the pluggable domain-speci fic layer that is adaptable to domain-specifi c applications. The multiple layers of LinkedMDR ensure its genericity and extensibility.

 
The overall Model Architecture of LinkedMDR

Hereunder, we provide links to documents and web pages related to our study: