Managing highly correlated semi-structured data: architectural aspects of a digital archive

XML techniques are well suited to describe, manage, store, and exchange hierarchical, semi-structured data. Information represented beyond hierarchical structures can still be described and exchanged in XML format employing additional concepts such as ID/IDREF or XLink. However, retrieval, manipulation, and storage mechanisms are far away from being the ideal solution for such data. Query languages do not perform efficiently in these cases. Especially in scenarios, such as the Digital Wossidlo Archive (WossiDiA), a project dealing with a huge number of arbitrarily correlated data units, XML query evaluation and retrieval techniques face problems, such as intricate querying and bad efficiency. At this point a solution to manage these data efficiently needs to be devised. This paper introduces a first approach which attempts to find such a solution for the WossiDiA information system.