MAPS: Manuscript map Annotation and Presentation System: Linking formal ontologies with social tagging to (re-)construct relationships between manuscript maps and contextual documents

Introduction The proposed paper describes a pilot entitled MAPS
of the Virtual Knowledge Studio for the Humanities
and Social Sciences (VKS) of the Royal Netherlands
Academy of Arts and Sciences (KNAW), the Dutch
National Archives, Leiden University Library and the
Department of Information and Computing Sciences of
Utrecht University intended to disclose manuscript map
collections in the Netherlands and potentially abroad
for individual and institutional research in the humanities.
The research is embedded in the Spatial/GIS lab
of a digital humanities collaboration of institutes of the
Royal Academy, the so-called “Alfa-Lab” that started in
March 2009.
It builds upon research and existing “draw over image”
software, developed within the context of the Paper and
Virtual Cities project, a collaboration between the University
Groningen and the Virtual Knowledge Studio.
MAPS provides a system in which users reconstruct historical
contexts by means of annotations and bottom-up
geo-references, here called “spatial tags” complementary
to the formal ontology based Encoded Archival Description
standard. In a technological sense this implies the
interoperability between (semi-)structured and “fuzzy”
annotations in the form of tagged draw-over-images and
the (semi-)automatic linking of maps with contextual
documents. Besides technical issues, methodological
questions are discussed regarding the role and authority
of annotations and tags of professional and non-professional
humanities researchers in a Web 2.0 environment.
The requirements for use of manuscript maps in humanities
research are mapped by means of observational user
studies of annotation practices and by testing interfaces.
Manuscript Maps
After the Fall of Antwerp (1585), Amsterdam, with famous
map printers/ publishers such as Blaeu, became in the 17th Century a leading cartographical center in the
world. Research of this predominant position in commercial
cartography is relevant for historians of cartography,
for historians of the book and publishing and for
cultural studies, such as art history. However, this emphasis
on printed commercial cartography also blurs the
importance of manuscript maps for research in a much
wider field of individual historians or of cultural heritage,
planning and policy institutions that use historical
maps for technical or administrative purposes. Manuscript
maps are significant not only for their quantity,
the Dutch National Archives alone count approximately
300.000 manuscript maps (including drawings) from
Dutch administrative bodies, but also for their quality
as unique historical sources. (De Vries 1989, Zandvliet
1998, Heuvel 2003 and 2004). Unfortunately, manuscript
maps are less accessible; their proper meaning can
only be understood when they are re-linked to the documents
they originally were intended to illustrate. Due
to 19th and 20th century archival processes many maps
lost their contexts when they were separated from the
documents to which they belonged. Encoded Archival
Description (EAD) and its extension Encoded Archival
Context (EAC), used as standards in many Dutch
archives, libraries and museums (including the Dutch
National Archives), are powerful description and contextualization
formats to encode links between maps and
documents dispersed over cultural heritage institutions.
Although the first experiments (coordinated by the Dutch
National Archives for the Netherlands and with Leiden
University Library as participant) with such formats are
promising and will be fully exploited, this linking process
to reconstruct original contexts and to create new
ones is time-consuming and cannot be done by cultural
heritage institutions on their own. A more productive
way is to involve so many users as possible in the (re)
establishment of these historical relationships by enriching
these manuscript maps with annotations, and to link
these to archival documents as described in inventories.
Annotation
In order to identify maps and to use them for historical
research, it is important to link them to geographical
space. Geo-referencing allows combinations and
overlays of maps with minimum distortion. However,
this process is quite difficult for manuscript maps that
vary so much in scale, precision, color etc. Instead of
complex geo-referencing systems we opt for existing
draw-over-image software, developed within the context
of the Paper and Virtual Cities project, a collaboration
between the University Groningen and the Virtual
Knowledge Studio that allows us to select any object on
the map for annotation in XML. (Heuvel & Koster 2007,
Benavides& Heuvel 2008), On the one hand the map
annotator module will be used by individual experts in a
conventional way to add knowledge through structured
annotations based on professional research. Leiden University
Library is a research library, whose special collections
are used for the greater part by scholars. On the
other hand much more fuzzy and fragmented knowledge
is available from users, who work more incidentally with
maps, e.g. genealogists, local historians and private collectors.
Their input through Web 2.0-techniques may be
of great help to reconstitute the lost context of historical
maps. They provide the required critical mass; 85% of
the visitors to the Dutch National Archives are genealogists.
However, we have to learn from Web 2.0 failures
(Crotty 2008) and avoid mismatches between tools and
the researchers that use them. Therefore analysis of annotation
practices and usability testing of interfaces with
focus groups of researchers in the humanities makes part
of the project. The methods used by E. Heere (2008) for
analyzing the use of a historical GIS interface, will be
extended to the analysis of practices of annotating/tagging
by researchers in a Web 2.0 environment. Another
issue that will be addressed is the authority of annotations
by expert and lay-experts. Hereto we try to build on
existing software (such and GENTECH and HarvANA)
that allow expressing and visualizing the provenance and
levels of authority of certain assertions. (Hunter 2008)
Presentation
Users have an easy access to the archive inventories
described in EAD in order to search and annotate digitized
manuscript maps and related sources. Annotations
of geographical space and time are described in specific
XML vocabularies, such as Geographical Markup Language
and Historical Event Markup Language. After
linking these clusters of knowledge have to be made
available to researchers according to their needs. This is
done in mash-ups maps (Google Maps, using the KML
standard). Because map annotation will be an ongoing
process, the user interface should be integrated in a coherent
system offering not only reference lists (tag lists,
gazetteer and archive inventories), but also the already
tagged documents.
System
MAPS enhances the research infrastructure by building
an annotation infrastructure for user generated research
content in the humanities, by geo-semantic data
integration (in this case maps, annotations and archival
data), and by the presentation in suitable user-interfaces
for researchers in the form of mash-ups. (Ennals 2007)
The geo-semantic data integration between maps, annotations
and archival data will be based on (semi-) automatic
linking, using multi-agent technology, between profiles of (historical) archive creators, typologies of
maps and profiles of individual experts and lay expert
researchers and research groups and cultural heritage,
planning and policy institutions that use historical maps
for technical or administrative purposes. The Dutch researcher Peter Sigmond studies Dutch
habours between 1500-1800 for his PhD research. In
chapter of his dissertation with the title Nederlandse
Zeehavens tussen 1500-1800 he discusses a report and
some designs of an engineer Alberdingh of 1688 for the
creation of dock harbors in several Dutch harbors as
alternative to the marine one in Amsterdam that gradually
silted up and was not accessible anymore for heavy
warships. Sigmond ordered a photograph of a part of a
manuscript map number 4 of one of these designs for
the town of Willemstad in the collection of Leiden University
Library (LU). In our scenario this library points
him to the draw-over-image tools [1] that allows him to
indicate the exact contours of the object on the digital
map in its image repository, that he wants to use for his
illustration. He sends this image with the tag dock to
LU [2]. His PhD also contains a reference to a report
in the Dutch National Archives (NA) that describes the
pro and cons of the various harbor cities for which the
designs were made. He tags this information to the drawover-
image (DOI) and sends it to LU [3]. DOI and tag
are linked to the metadata description of the map 4 in
EAD that has already a link to three others files describing
maps belonging to the same project [4]. Numbering
of the maps also make clear that there are some missing
maps. Since the report refers to the Dutch National
Archives a link is made to sources described with EAD
metadata of this archive, including the DOI and tags [5]
and reconnected with the maps of LU [6]. In a query on
expected archive creator profile, in this case military, the
database shows that there is another design of the same
year for the same project map 5 and another later copy of
it. The image of LU and the metadata of all their relevant
images are combined with those of NA[7] and send with
the photographical order to the user [8]. With this information
the series of designs in Leiden University Library
is completed with one extra drawing from the NA. The
PhD researcher found out that there were extra drawings.
Finally the National Archive found that there were more
drawings belonging to the report in their collection.
References
Benavides J. and Heuvel, C. van den (2008). The function
and accuracy of old Dutch urban designs and maps.
A computer assisted analysis of the extension of Leiden
(1611) in Digital Humanities 2008 Conference, University
of Oulu, Finland June 24-29, 55-56.
Crotty, D. (2008). Why Web 2.0 is failing in biology.
Harbor Protocols Cold Spring http://www.cshblogs.
org/cshprotocols/2008/02/14/why-web-20-is-failing-inbiology/
EAD/EAC. http://www.iath.virginia.edu/eac.
ECAI. http://www.ecai.org/
Ennals, R. and Gay, D. (2007). User-friendly functional
programming for web mashups. In Proceedings of the
2007 ACM SIGPLAN international Conference on Functional
Programming (Freiburg, Germany, October 01 -
03, 2007). ICFP ‘07. ACM, New York, NY, 223-234.
GENTECH: http://www.ngsgenealogy.org/ngsgentech/
projects/Gdm/Gdm.cfm.
Heere, E. (2008). GIS voor historisch landschapsonderzoek:
opzet en gebruik van een historisch GIS voor
prekadastrale kaarten, PhD thesis Utrecht University 27
June 2008.
Heuvel, C. van den (2003) “Atlasticity”, Problems in
defining and the digitizing military manuscript atlases of
the Low Countries. In [ed. I. Warmoes, E. d’Orgeix, C.
van den Heuvel] Atlas Militaires Manuscrits Européens
(XVIe-XVIIIe siècles). Forme, contenu, contexte de réalisation
et vocations, Actes des 4es journées d’ étude du
Musée des Plans-reliefs, Paris (18-19 April 2002) Paris
pp. 11-26.
Heuvel, C. van den (2004). Mapping Mixed Maps. Historical
and future constructions of time and space in urban
cartography. In Le temps 129e congrès des sociétés historiques et scientifiques, Colloque IV Le temps des
cartes. Monde des cartes, Bulletin du Comité français de
cartographie [Numéro special],pp. 23-40.
Heuvel, C. van den, Koster E. (2007). Tussen papieren
en digitale kaarten. De annotatie van een atlas en bijbehorende
kaarten van Woerden. In Aangeraakt. Boeken in
contact met hun lezers. Een bundel opstellen voor Wim
Gerritsen en Paul Hoftijzer, red. Kasper van Ommen,
Arnoud Vrolijk, Geert Warnar, Kleine publicaties van de
Leidse Universiteitsbibliotheek Nr. 75, pp. 150-156.
Hunter, J., Khan, I., and Gerber, A. (2008) HarvANA
-Harvesting Community Tags to Enrich Collection
Metadata. In ACM IEEE Joint Conference on Digital Libraries,
JCDL June 16 - 20, 2008. Pittsburgh.
Vries, D. de [ ed.] (1989). Kaarten met geschiedenis
1550-1800. Een selectie van oude getekende kaarten
van Nederland uit de Collectie Bodel Nijenhuis HES,
Utrecht
Zandvliet, K. (1998). Mapping for Money. Maps, plans
and topographic paintings and their role in Dutch overseas
expansion during the 16th and 17th centuries, Batavian
Lion international, Amsterdam.