The World Wide Web (WWW), as an ubiquitous medium for publication and exchange, already significantly influenced the way how historians work: the availability of public catalogs and bibliographies enable e cient research of relevant content for a certain investigation; the increasing digitization of works from historical archives and libraries, in addition, enables historians to directly access historical sources remotely. The capabilities of the WWW as a medium for collaboration, however, are only starting to be explored. Many historical questions are only answerable by combining information from di erent sources, from di erent researchers and organizations. Furthermore, after analyzing original sources, the derived information is often more comprehensive than can be captured by simple keyword indexing. In [3] we report about the application of an adaptive, semantics-based knowledge engineering approach for the development of a prosopographical knowledge base. In this demonstration we will showcase the comprehensive prosopographical knowledge base and its potential for applications. In prosopographical research, historians analyze common characteristics of historical groups by studying statistically relevant quantities of individual biographies. Untraceable periods of biographies can be determined on the basis of such accomplished analyses in combination with statistically examinations as well as patterns of relationships between individuals and their activities. In our case, researchers from the Historical Seminar at the University of Leipzig aimed at creating a prosopographical knowledge base about the life and work of professors in the 600 years history of University of Leipzig ranging from the year 1409 till 2009 - the Catalogus Professorum Lipsiensis (CPL).
Linked Data
Partial RDF export synchronize
Model Data (SPARUL) Linked Data
Full RDF export backup Model general web user experienced
web user content editor (Project Team) browse, search query, search
OCPY ue r g cocnofinqfiguuerery, seagrecthData add, edit, maintain [stable]
[experimental]
synchronize
Model Data (SPARUL)
browse, annotate, discussHTML GUI
The semantic data wiki OntoWiki [
For general web users the catalog is integrated in the public website of the University of Leipzig2. A simplified user interface consisting of plain HTML and Linked Data3 resources is generated nightly from the knowledge base. The historians are able to interact with CPL via an experimental version4 of OntoWiki. The version of the catalog available there is synchronized with the protected OntoWiki installation, transforms the exported data considering any linked knowledge bases and imports the changed data into this experimental installation. The project platform is usable for humans accessing web interfaces for reading or editing data. OntoWiki also provides a number of generic access interfaces, which include SPARQL and Linked Data endpoint and a Semantic Pingback server. On top of these generic access interfaces application specific access interfaces are deployed. These are described in more detail in the remainder and will be presented in the demonstration. 1 Available at: http://catalogus-professorum.org/cpm/ 2 http://www.uni-leipzig.de/unigeschichte/professorenkatalog/ 3 http://www.w3.org/DesignIssues/LinkedData.html 4 http://catalogus-professorum.org/
Public CPL website and Linked Data CPL is not just a tool for historians, but aims to showcase the results of historic research to the wider general public. For that purpose a special public website was created. The user interface of the public website is geared towards simplicity. The knowledge base can be explored by epochs, faculties, functions of professors (i.e. rector or dean) or alphabetically. Professors of the day are automatically selected based on important days in the life of a professor (i.e. birth or death). Furthermore, the public website comprises a full-text search, which searches within all literals stored in the CPL knowledge base.
Using OntoWiki’s build-in endpoint functionality CPL is immediately available as Linked Data. Within the Linking Open Data e ort, hundreds of data sets have already been connected to each other via owl:sameAs links. By interlinking CPL with other related datasets, we aim at establishing CPL as a linked data crystallization point for academic prosopographical knowledge. 4
OntoWiki also serves as a SPARQL endpoint, however, it quickly turned out that formulating SPARQL queries is too tedious for the historian domain experts. In order to simplify the creation of queries for the historians, we developed the Visual Query Builder5 (VQB) as an OntoWiki extension, which is implemented in JavaScript and communicates with the triple store using the SPARQL language and protocol. VQB allows to visually create queries to the stored knowledge base and supports domain experts with an intuitive visual representation of query and data. Developed queries can be stored and added via drag-and-drop to the current query. This enables the reuse of existing queries as building blocks for more complex ones. VQB also supports the set-based browsing paradigm by visualizing di erent connectives, such as join, union, 5 http://aksw.org/Projects/OntoWiki/Extension/VQB intersection, di erence between queries. The incremental query building is facilitated by displaying results already during query creation. The VQB user interface is visualized in Figure 2. The user interface can be adjusted by scaling or deactivating unused panels. 5
An important aspect of historical investigations is the search for relationships
between di erent persons or entities of interest. An application supporting such
investigations within RDF knowledge knowledge bases is RelFinder6 [