Archivkopie der Website http://www.humboldt-im-netz.de |
---|
______________________________________________________
![]()
Detlev Doherr
The Humboldt Digital Library: Exploring Innovative Structures
3. Project Developments and Testbed
In order to demonstrate the functionality of this information system, we developed a prototype that contains some basic functions for text search within all documents, tools to access volumes, chapters and paragraphs, and a search method for keywords or related synonyms. To avoid critical data transformation from XML and SQL structures we provide the page-oriented scans for each paragraph or footnote. Those page numbers are given in parentheses and are linked with the image itself. Thus, the user may click on the marked page numbers in the digital text to see the scan of the original text or image.
The digital information system is based on modern client-server technology. This system contains the complex SQL structures of our database with Humboldt’s texts and images, structured in paragraphs. This solution provides many more innovative features than the traditional digital library:
Web services for information, not for documents
Comparative analysis of paragraphs in different languages
Identification of keywords and synonyms for multilingual approach
Thematic search
Chronology of travels
Identification of names
Links to archives from particular text paragraphs
For the presentation of geographic or thematic maps we need an interactive map server on the Internet, and this map server must contain digitalized information and provide a link to the text archive. We developed access to texts with points of interest on Humboldt’s Orinoco map and tested navigation features. Our online prototype shows a simple Orinoco map which is georeferenced and has some marked points in thematic layers (See fig. 1
LINK). Our experiment has been successful and shows the usefulness of navigation from texts to maps and their thematic layers.
Interactive map (Humboldt’s travel map to the Orinoco) with interactive layer and points of interest, providing access to the original texts.
Furthermore, we also investigated the problem of accessing texts depending on a location on the map. We developed a search function to use a keyword, which is related to the entire point of interest in the interactive map. We defined objects on the map, which have attributes such as themes, dates, or notes. This problem revealed the usefulness of SQL structures for the correlation of objects inside the maps and the harvesting of information from the digital archive. This is precisely the structure that we have developed successfully in our prototype.
Unfortunately, SQL structures involve a complex database design. So we defined media assets, which contain the formatted text, images, scans or maps. In addition, the digitized and formatless texts and notes are part of those assets. Together with unique identifiers for paragraphs, related keywords, themes, synonyms and other elements, it is possible to create dynamic links to other assets, no matter if maps, text or images. The system will provide access to those assets via Web browser. The linking can be predefined in expert shells, but also created automatically, whenever global information is necessary.
______________________________________________________
<< letzte Seite | Übersicht | nächste Seite >>