That obscure object of desire: multimedia metadata on the web
Frank Nack, Jacco van Ossenbruggen, Lynda Hardman · Centrum Wiskunde & Informatica (CWI), the national research institute for mathematics and computer science in the Netherlands · 2003
In this article we discuss the advances in, and remaining problems of, making use of audio-visual media in a semantic-based environment, such as the Semantic Web, facilitated through media-aware and ontologybased metadata.Our discussion is predominantly motivated by the two most widely known approaches towards machine-processable and semantic-based content description, namely the Semantic Web activity of the W3C [2, 4] and ISO's efforts in the direction of complex media content modeling, in particular the the Multimedia Content Description Interface (MPEG-7) [30, 31, 32, 33, 34].We chose these two approaches as they provide the potential techniques to establish a media-aware Semantic Web, even though at the time of writing the approaches seem to be diverging rather than converging.The Semantic Web should bring machine-processable content to Web pages, thus being an extension of the current Web.The aim is to add ontology-based metadata to Web resources to improve Internet search and provide means for machine-based reasoning about the content.A major drawback of the current semantic Web developments, however, is its media-agnostic view on Web resources.The specific needs of dynamic audio-visual media with its variety of data representations is not recognized.That is, however, precisely what the intention of MPEG is, partially addressed in MPEG-4 and fully in MPEG-7.In this paper, we explain that the conceptual ideas and technologies discussed in both approaches are essential for the next step in Web-based multimedia development.Unfortunately, there are still many practical obstacles that block their widespread use for providing multimedia metadata on the Web.We show that a media-ware Semantic Web will blur the boundaries between traditional categories like preproduction, production, and postproduction, with far-reaching effects on concepts such as data, metadata, consumer and producer.The paper is structured as follows.We first provide a scenario to explain our vision of a media-aware Semantic Web and derive from it a number of problems regarding the semantic content description of media units.We then discuss the multimedia production chain, in particular emphasizing the role of progressive metadata production.As a result we distill a set of media-based metadata production requirements and show how current media production environments fail to address these.We then introduce those parts of the W3C and ISO standardization works that are relevant to our discussion.We analyze their abilities to define structures for describing media semantics, discuss syntactic and semantic problems, ontological problems for media semantics, and the problems of applying the theoretical concepts to real world problems. Metadata in the multimedia production chainAudio-visual (AV) media production, such as for news, documentaries, interactive games, virtual environments, or business presentations as described in the scenario, is a complex process, that differs from text production in many aspects.An author of a text, who wants to communicate a message, can state this message explicitly in the text.A producer of a piece of AV material, however, usually needs to package the message implicitly within the AV content.Therefore, creating metadata that makes this type of communication explicit is even harder for AV material than it is for text.Another issue that complicated media annotation is defining the part of the AV content that the annotation is about.For text, these techniques for pointing into a larger text are well established, both in