Time: Thursday, May 10, 2007 (3:30pm-5:00pm)
Location: New Brunswick
Organizers and Moderators:
The panel is proposed by the W3C Multimedia Semantics Incubator Group, and supported by the European Commission under contract FP6-027026, Knowledge Space of semantic inference for automatic annotation and retrieval of multimedia content - K-Space.
Are multimedia metadata really needed on the Semantic Web? Are "real" multimedia standards, such as MPEG-7, of any value at all to the Web community? Do we really need to invoke the complexity of formal knowledge representation languages and ontologies?
The Web already supports the storage and retrieval of media such as text, image, video and audio. In contrast to text, other media types require some sort of metadata in order to make them indexable and searchable. Surrounding text on the web page and user assigned tags are current methods which give access to millions of images and videos to millions of users. The Web 2.0 paradigm has become successful partly because it asks almost nothing from its end users. So, is there even a problem to be solved?
Existing multimedia metadata standards, such as MPEG-7, provide a means of associating semantics with particular sections of audio-visual material. Expensive and complex tools are available for giving professional users access to this. Many other metadata standards, such as EXIF, ID3 and XMP, also exist, making it difficult for a typical Web user to see the wood for the trees.
The panel will discuss how to overcome the shortcomings of current mulimedia metadata formats. In addition, the Web community needs to understand the benefits of explicit semantics for retrieving and process multimedia content. We will discuss whether standards for multimedia metadata for Web content are really needed, and what role they could play in a future semantic multimedia web.
In particular, we look for answers to the following questions: