=Paper=
{{Paper
|id=None
|storemode=property
|title=Prototyping a Browser for a Listed Buildings Database with Semantic MediaWiki 
|pdfUrl=https://ceur-ws.org/Vol-632/paper14.pdf
|volume=Vol-632
|dblpUrl=https://dblp.org/rec/conf/semwiki/Dumitrache0KA10
}}
==Prototyping a Browser for a Listed Buildings Database with Semantic MediaWiki ==
<pdf width="1500px">https://ceur-ws.org/Vol-632/paper14.pdf</pdf>
<pre>
    Prototyping a Browser for a Listed Buildings
        Database? with Semantic MediaWiki

    Anca Dumitrache1 and Christoph Lange1 and Michael Kohlhase1 , and Nils
                                Aschenbeck2
             1
                Computer Science, Jacobs University Bremen, Germany
           {a.dumitrache,ch.lange,m.kohlhase}@jacobs-university.de
              2
                aschenbeck media UG, Bremen, info@aschenbeck.net



1     BauDenkMalNetz and its Intended Applications

Listed buildings, even if they are not top landmarks, are increasingly attracting
visitors. People express interest in hidden gems in their neighborhood or along
their travel itinerary, and in the history of the building they live in. All required
data has been meticulously collected by the offices for historical monuments but is
not flexibly accessible. In Bremen, the database of buildings (with location, map
of the estate, construction history, architect, photos) is searchable and browsable
online3 , but that only helps users who know how to use a rigidly structured
database search form. Our beginning BauDenkMalNetz effort (“listed building
web”) aims at a wider purpose: modeling the semantic structure of these data,
starting in Bremen but open for other data, and exposing them via a semantic web
interface with enhanced querying and presentation capabilities. Requirements
beyond interactive browsing comprise auto-generation of customized printed
guides (e. g. “Bauhaus villas in my neighborhood”), on-demand presentation on
mobile devices (e. g. “medieval churches along my travel itinerary”), and serving
linked data for usage by other services.


2     Exploring Possibilities in Semantic MediaWiki

These requirements clearly demand semantic technologies. In this early phase,
the possibilities of how to represent our knowledge, how to utilize it, and how to
represent it to users are not yet entirely clear to us. Therefore, we have started
building a prototype using Semantic MediaWiki (SMW [4]). Thanks to its stable
MediaWiki foundation, its customizability and the wide availability of extensions,
SMW is a preferred choice for building prototypes (see, e. g., [2]). In our case,
another motivation is that it has already been used for conceptually similar
applications. The Archiplanet [1] SMW site contains over 100,000 pages about
buildings and architects, however, with a semantically rather weak ontology. We
?
  We thank the Landesdenkmalamt (state office for historical monuments) Bremen for
  their data, and Axel Polleres and the SMW community for technological advice.
3
  http://194.95.254.61/denkmalpflege/
are instead planning to follow the existing, stronger database schema of our
data for incrementally developing an initial ontology, which is easy in SMW. We
have started manually annotating a small, strongly interlinked dataset of listed
buildings in one district of Bremen in that way4 and will explore first possibilities
for services on these data, drawing on the abundance of available SMW extensions:
Semantic Forms is an extension for form/template-based user interfaces, providing
forms for adding, editing and querying data, that allows for complex in-document
annotations, like embedding forms into other forms by using form templates. This
feature will allow for representing fine-granular entities, which nevertheless have
some properties of interest, as annotated fragments of larger wiki pages, instead
of having very small pages created for them. The custom templates created with
Semantic Forms, which are adapted to the structure of the data, will also help us
in a later step of the project, when we will automatically import a large amount
of data entries from the existing relational database by a bot. Semantic Drilldown
allows for interactively drilling down through different dimensions (= properties)
of data. The whole range of values and the number of values is visible from
the beginning. This extension enables filtering by semantic properties based on:
property values (e. g. the address of a building), page categories (e. g. the district
to which a building belongs), date ranges (e. g. the years between which the
house was built). Maps and Semantic Maps are extensions for integrating Bing
and Google maps; we will display the locations of buildings on maps. Semantic
Maps supports compound queries, like which category an item belongs to, that
also work with geo-coordinates. Semantic Graph is an extension for visually
representing results of complex queries as graphs. That can serve to illustrate
relations inside the ontology, like “part of” (one building being part of another
building) or “time” (by creating a chronological alignment of buildings).


3     Conclusion and Outlook
Using SMW and its extensions, we are creating a functional prototype of the
“listed building web”, which will be expanded into a full-fledged web portal. We
are planning to enhance our initial project-specific ontology by reusing CIDOC
CRM, a standard ontology for cultural heritage [5], and GeoNames, a standard
ontology for geospatial information, for which a number of web services exist [3].


References
[1] Archiplanet. url: http://www.archiplanet.org.
[2] Jie Bao, Li Ding, Rui Huang, et al. “A Semantic Wiki based Light-Weight
    Web Application Model”. In: ASWC. 2009.
[3] GeoNames. url: http://www.geonames.org.
[4] Semantic MediaWiki. url: http://semantic-mediawiki.org.
[5] CIDOC Conceptual Reference Model. url: http://cidoc.ics.forth.gr.

4
    See http://mathweb.org/wiki/BauDenkMalNetz

</pre>