XML is the standard format for data exchange over the internet and the Web contains millions of XML documents. Even though the structure of XML documents can be very flexible, it is desirable for many applications that documents come with a schema that describe their structure in a concise way. Whereas there exists a lot of software for the manipulation of XML documents and modern database management systems can deal with XML, the foundations of XML schema and document management systems are still not fully understood and they constitute an active research area.
Schema languages for XML, most prominently XML Schema and DTD, are closely related to concepts from Formal Language Theory, including context-free grammars, tree automata and regular expressions. However, the official standards of the
World Wide Web Consortium
(W3C) pose various restrictions that have not been much studied in Formal Language Theory before the advent of XML (and its precursor SGML). Altogether, the foundations of XML schema and document management systems raise a lot of new challenges to Automata Theory, in particular, to provide suitable algorithms and concepts.
The aim of this talk is to describe some of these challenges, to report on recent developments and to highlight some current directions of research. Topics will include the expressive power of schema languages, the repair of schemas that do not obey the restrictions posed by the W3C standards, the combination of schemas, schema conversion, schema minimisation, inference of schemas from a given set of documents, segmentation of schemas for distributed documents, and constraints that combine structural and data aspects.