Information Technology Reference
In-Depth Information
information. By using hCard, we can explicitly indicate telephone numbers like
<span class="tel">0678901234</span> . Microformats also enable us to pre-
serve expressiveness of HTML itself and to optionally introduce semantics of
data in HTML documents.
From these observations, we believe that there can be various useful applica-
tions of optional type annotation for webbles as well. However, we have to keep
in mind that webbles is not a RAD (Rapid Application Development) tool. It is
a media architecture. Therefore, we should not limit possible annotation types
for webbles only to those used for data types in programming.
For instance, if you make a dialog webble for entering book reference data,
it is a good idea to annotate both the webble and its slots, using the Dublin
Core vocabulary. In this way, if later on you develop a new application, say for
EPUB topics, you can easily find that dialog from a webble repository and al-
most automatically connect it to your new application. This is possible because
Dublin Core is not only a programming technology but also a common vocab-
ulary used for annotating files of various media formats, including EPUB. We
can also imagine similar examples not only for Dublin Core, but also DBpedia,
YAGO, AAT, MeSH and any other controlled vocabularies used in the Web. By
annotating webbles with such widely used vocabularies, webbles can gradually
become citizens of the Linked Open Data (LOD) world.
In this paper, we introduce a generic and formal framework for annotating
webbles. The kind of vocabulary that we envisage can be a pre-defined vocabu-
lary specific for the webble framework, or the LOD world. The framework that
we propose here works with any kind of controlled vocabulary. Moreover, our
framework doesn't touch the webble architecture itself - it just overlays an ad-
ditional, semantic layer on top of the naked webble architecture.
2 Webble Annotations
The current webble technology offers some means to attach annotations both at
webble level and at slot level. At webble level we find several kinds of annota-
tions in the current webble implementation: name, class name, developper, group
identifiers, description keywords and “metadata”. However, the terms in these
annotations do not come from any standard vocabulary and the mechanism for
creating metadata is rarely used (i.e. most webbles have empty metadata).
At slot level, we find two kinds of annotation: (a) a slot name and (b) a
slot type. However, none of these information items is exploited from the user's
point of view. Moreover, there is no mechanism available to express attributes
(or properties) of a slot other than its name and type.
Another problem is that the names in these annotations do not come from
any standard vocabulary either. For instance, if someone wants to search webbles
by name, he has to read a document about webble file format to identify the
name of the annotation which represents webble names. This lack of relationship
with widely-used controlled vocabularies prevents visibility, searchability, and
reusability of webbles in the context of the world-wide-web.
Search WWH ::




Custom Search