---

The Web is continuously being transformed from a web primarily aimed at human consumption to a web of data which will allow autonomous agents to carry out complex reasoning tasks for humans. An indispensable ingredient for this transformation is rule based data integration of semi-structured web content in its various formats, such as HTML, XML, RDF and Microformats. By introducing the framework of Rich Unification, this thesis shows how existing rule languages can be adapted to fulfill the needs for data integration on the Web. It is shown that SPARQL, XPath and Xcerpt neatly fit into this framework. Xcerpt is extended to become a truly format-versatile web query language within the bounds of rich unification. One of these extensions is the expressive, yet polynomial time RDF path query language RPL. A model theoretic semantics for RDF containers, RDF collections and reification in the style of the RDF/S semantics is introduced as the formal foundation of query constructs specifically aimed at querying these RDF specifities. Both RPL and the other RDF query constructs are amenable for embedding in other RDF rule languages such as SPARQL or Triple.

---

Benedikt Linse has received his PhD in computer science from theUniversity of Munich in 2010. He has contributed to numerousinternational conferences and books on rule languages, SemanticWeb querying and Web technology.

---