---

The project is aimed to develop an XML update language built on top of XQuery by adding on update primitives, which complies with the standard XQuery Update Facility 1.0 requirements. The end product does this in a flexible way, using a software archi- tecture which enables developers to choose among different XQuery engines and various XML Processing APIs in order to achieve the overall optimal performance for queries with different characteristics. This thesis starts with an investigation on existing proposals and implementations for XML updates, and describes the approach for the proposed XML update language, followed by an explanation of its implementation. In the end three groups of experiments are undertaken to evaluate the performance of this implementation: software- related, query-related, and tests for other factors. Software-related experiments identify the overall optimal implementation of the software. Query- related analysis studies how the performance varies for queries with different characteristics. Tests for other factors investigate the limitations of the software (i.e. input file size and disk space requirement).

---

Yuan Tian works for Goldman Sachs in London and holds an MSc degree in Computer Science from the University of Oxford.

---