XML query processing: efficiency and optimality

XML (Extensible Mark-up Language) is a well established format which is often used for modeling of semi-structured data. XPath and XQuery are de facto standards among XML query languages and searching for occurrences of a twig pattern query (TPQ) in an XML document is one of their core tasks. There is a large number of different approaches addressing the TPQ matching problem. The aim of this article is to compare the state-of-the-art techniques and give an overview which can help to understand the relationships between different methodologies used in this area. We distinguish three main areas of a TPQ processing: (1) index data structures and XML document partitioning, (2) join algorithms, and (3) cost-based optimizations. We cover the most important techniques in each area and explain their relationships and possible combinations.