This XML retrieval paradigm implies to change the way systems are evaluated. In INEX (see below), a new assessment scale has been proposed along with new precision/recall metrics (the Norbert Gövert PRng metric for instance). I proposed a  several metrics, the latest being the most expressive (generalisation of precision-recall) and simple to compute  (missing reference).

I also co-authored for the Encyclopedia of Database Systems a paper on XML evaluation metrics (missing reference).