Domain Knowledge Representation in Semantic MPEG‐7 Descriptions

Semantic‐based multimedia retrieval and filtering services have recently become very popular. This is due to the large amount of digital multimedia content that is produced everyday and the need for locating, within the available content, the multimedia content that is semantically closer to the preferences of the users. Fortunately, the dominant standard for audiovisual content description today (MPEG‐7) allows for the structured description of the multimedia content semantics.In addition, the use of domain knowledge in semantic audiovisual content descriptions enhances the functionality and effectiveness of the multimedia applications. However, the MPEG‐7 does not describe a formal mechanism for the systematic integration of domain knowledge and reasoning capabilities in the MPEG‐7 descriptions. The specification of a formal model for domain knowledge representation and reasoning using the MPEG‐7 constructs is of paramount importance for exploiting domain knowledge in order to perform semantic processing of the multimedia content. In this chapter, we present a formal model that allows the systematic representation of domain knowledge using MPEG‐7 constructs and its exploitation in reasoning. The formal model that we describe exploits exclusively MPEG‐7 constructs, and the descriptions that are structured according to the model are completely within the MPEG‐7 standard.