XML makes it possible to describe such data in a single
document, and it provides a set of tools to deal with them.
In the next section we will present what can be done with
these documents.
5. PUBLISHING DATA WITH XML
In the context of this application our objective is to
publish the data in different ways.
5.1 A single data source for multiple target
Many different users (from simple visitors to domain
expert) will have access to published data. We have to
provide them with a view of the data adapted to their
needs, that is to say we have to build a dynamic
presentation of the data (for example a web-site)
depending on the profile of the user.
The presentation also depends on the kind of client:
browsing from a workstation, from a handheld device or
preparing a printable document.
Presentation
Dynamic HTML Dynamic XHTML PDF
Web sites + JPEG images | | Web sites + JPEG images | | documents
Transformation
XSL transformation
ES
Data Sources
r- Virtual dynamic XML Documents
L— -XMI Documents Database
Figure 5. XML publishing framework
5.2 XML publishing framework
XML provide a unified framework to model the data, to
transform them, and to present them in a dynamic web-
site or in high-quality printed documents. A set a tools
has been proposed (each is based on XML):
XSLT? (Extensible Stylesheet Language
Transformation), is a language dedicated to the
transformation of the tree structure of XML document.
We are using it to produce every target formats from the
sources.
XHTML/, is a modularized and XMLized version of
HTML. We are using it to provide high quality web sites
to various clients (from workstations to handheld
devices). A classical HTML version will be also provided
for backward compatibility.
SVG? (Scalable Vector Graphics), is a new generation
graphics format. As it is XML based it can be
dynamically produced from XML sources. We are using
it to produce dynamic maps or to enhance underwater
3 http://www.w3.org/TR/xslt
* http://www.w3.org/TR/xhtmll1
° http://Www.w3.org/TR/SVG
pictures with marks on detected objects (bounding boxes). It
can be integrated in XHTML documents to provide interactive
graphics (an easy way to browse the data). For older browser
those pictures can be dynamically transformed to a bitmap
format (the interactive aspect is lost).
X3D* (Extensible 3D), this format dedicated to XMLized 3D
modelling and adapted from VRML’97 is still under
development. We will use it to provide a dynamic 3D view
The whole system will then be XML based the data sources
will be dynamically transformed in a web site where the users
will browse: presented data, interactive pictures or a 3D
model. A high quality printable version of the site will also be
generated.
But that may not enough for advanced user. They may need to
express their own queries over the data (like in a classical
database system).
5.3 Querying XML data
Like SQL for relational database systems, a functional query
language is under development for XML, it is called XQuery.
The important thing is that if all the data are expressed in
XML, XQuery will be a uniform way to query them:
photogrammetric data, 3D structure or comments could then
used in the same query.
But even if this language will be easy to learn, users must be
provided with easier tools to query the data. It will also avoid
repetitive construction of complex queries. To do this we will
use a dynamic generator of HTML forms, developed for the
Multimedia Search Engine (MUSE’) project.
With those kind of tools advanced users will be able to define
their own view of the data.
6. CONCLUSIONS
This work associates several laboratories working in various
disciplines and nevertheless complementary. Teams from with
different backgrounds are currently working on the same tool,
with the same XML formalism, each every team having taken
a few steps towards the others to harmonize the lexicon and to
establish a common language.
The management of data, an omnipresent problem in
archaeology, is dealt with in two ways: the first one is purely
textual and the second is from based on the object geo-
referential point-of-view. These two approaches being are
accessible over on the Internet.
The use of a three-dimensional model as an interface to the
data formalized in XML allows the purely documentary data
(references, observations made during the excavation,
photographs) to be linked to a three-dimensional
representation of the object. This graphic expression of the
object relies on the data (position, orientation, dimensions) and
on the generic knowledge of the object (theoretical shape,
default values, relationships between diverse objects). The
three-dimensional model, generated by the system, shows the
generic model of the object, defined by the archaeologist,
measured by photogrammetry and thereby a relevant interface
between the user and the collected data.
° http://www.web3d.org/x3d.html
7 htp://sis.univ-tln.f/muse
—346—
st C8 6 Ro As MM
~ ™ a T1
