Event-condition-action rule languages over semistructured data

The increasing use of semistructured data technologies such as XML and RDF, in dynamic applications in areas such as data integration and e-commerce make it necessary to support reactivity over such data. Event-Condition-Action (ECA) rules are one of the technologies that can provide such functionality. In this thesis we investigate potential languages and supportive technologies that enable reactive functionality over both XML and RDF data using ECA rules. We first review the syntax and the execution semantics of an ECA language for XML data and conduct a study regarding its expressiveness. We also describe an implementation of this language that can be utilised to provide reactivity over XML files and we study its performance using analytical and experimental methods. We introduce a new language for supporting ECA rules over RDF data, called RDFTL, and define its semantics. We also describe the architecture and implementation of a system that provides reactive functionality over RDF data in a peer-to-peer environment using RDFTL. Issues relating to indexing and management of RDF data, indexing and management of RDFTL rules, and change detection are discussed. We develop an analytical model for analysing the behaviour of our RDFTL rule processing system using as the main performance criterion the time required to complete all rule processing after an initial top-level update. For the purposes