Some thoughts on the general interoperability framework issue & more specifically the Mulimedia Semantics Authoring UC

Dear all,


below you'll find some detailed info about our SPF format. After the intial
positioning (see also the attached paper), we clearly see a link with the
Authoring UC. Moreover, it might be possible to our opinion to come up with,
for example, a SMIF (Scalable Metadata Interoperability Format) framework,
based on our SPF framework (re-implemented using ontologies and discarding
the presentation layer) which might tighten the semantic interoperability
gap a bit.  Please, comment on either (SPF, SMIF) of our thoughts.


Scalable Presentation Format (SPF)
=========================

Content providers struggle with the huge heterogeneity in devices consuming
multimedia today. They want to distribute their content to as many devices
as possible, however due to the differences in device capabilities (e.g.,
screen size), different versions of the same content have to be created. To
realize this goal, scalable content can be used. Research concerning
resources (text, video, audio, and graphics) has been done in order to
create scalable resources. However these resources are usually embedded in
multimedia presentations and therefore presentations must also be scalable
to realize the full UMA experience. The SPF is a presentation format that
can be adapted according to a specific target device.

SPF model  is characterized by a Separation of Concerns (SoC) on multiple
levels:

*       presentation metadata (information about the presentation such as
layout, font, navigation) ßà resource metadata (information about the
resources such as height, width, bitrate)
*       within the presentation metadata, a separation is introduced between
layout, style, and interactivity
*       the actual presentation metadata uses the User Interface Markup
Language (UIML), which is a meta-language for all possible text-based
presentation languages. It makes use of a vocabulary to map the UIML code
fragments to the dedicated language construct of the underlying presentation
format.

The current SPF model is implemented by using standardized technologies such
as MPEG-21 Digital Item Declaration (DID), UIML, and XML Path (XPath)  (
Note that the SPF model is independent of the used technologies). MPEG-21
DID is used to describe the structure of the resources. It is also used as
container format for the presentation metadata, hereby managing different
descriptors (layout, style, interactivity). As indicated above, UIML is used
for the presentation metadata. Within the presentation metadata, resource
metadata can be inserted by using XPath.

The benefits of the SPF are the following:

*       Use of standardized technologies
*       SoC which introduces a high degree of reusability
*       Limited redundancy when multiple devices has to be supported
*       Strong support for templating (i.e., every presentation is a
template for a class of resources)


Relation with the use case: Semantics from Multimedia Authoring
===============================================

In this use case, the lack of a presentation language that offers solutions
for including metadata into the multimedia presentation is presented. The
SPF (in its current form) itself is not the solution for this problem;
however, it can be used as a meta-format for all possible presentation
languages. Hereby, it is possible to include metadata about the presentation
(such as tags, properties of the presentation, tags for the included
resources) with SPF. To be more specific, SPF can be seen as a container
that will be able to manage all kinds of metadata needed for this use case.
First, within the resource metadata, all kinds of information can be added
related to these resources (from low-level information to high-level tags).
Second, properties of the presentation can be added to the presentation
metadata (as a separate descriptor). Finally, after translating SPF
(selecting the proper descriptors, insertion and adaptation of resources,
joining UIML code) into a UIML document, this document can be extended with
the valuable metadata about the resulting presentation. This metadata will
be a combination of the metadata of the used resources and the metadata
about the properties of the presentation.

There is, of course, still a need for a standardized format for the
description of properties and additional information related to resources
and presentation. However, such a standardized format could easily be
included within SPF.  Moreover, it might be possible to our opinion to come
up with, for example, a SMIF (Scalable Metadata Interoperability Format)
framework, based on our SPF framework (re-implemented using ontologies and
discarding the presentation layer) which might tighten the semantic
interoperability gap a bit.  Attached you'll find a paper on SPF itself.
Please, comment on either (SPF, SMIF) of our thoughts.



Sincere greetings,



Erik Mannens & Davy Van Deurzen

Ghent University - IBBT
Faculty of Engineering
Department of Electronics and Information Systems
Multimedia Lab
Gaston Crommenlaan 8 bus 201, B-9050 Ledeberg-Ghent, Belgium

t: +32 9 33 14993
f: +32 9 33 14896
m: +32 473 27 44 17
t sec: +32 9 33 14911
e: erik.mannens@ugent.be < <mailto:erik.mannens@ugent.be>
mailto:erik.mannens@ugent.be>

URL:  <http://multimedialab.elis.ugent.be>
http://multimedialab.elis.ugent.be < <http://multimedialab.elis.ugent.be/>
http://multimedialab.elis.ugent.be/>