TWIX: Approximate and Exact Twig Structure and Content Matching over XML Document Collections using Binary Labeling

XML queries specify predicates on the content and the structure of the elements of tree-structured XML docu- ments. Hence, discovering the occurrences of twig (tree structure) query patterns is a core operation for XML query processing. Prior works have typically applied top-down decomposition of the twig patterns into (i) binary (parent-child or ancestor-descendant) relation- ships, or (ii) path expression queries, followed by a join operation to reconstruct matched twig patterns. How- ever, most of these methods (i) rely on the user's knowl- edge of the underlying database to pose well-formed queries, and (ii) suffer from inspecting too many irrel- evant results. In this paper, we propose a novel heuris- tic for matching of XML twig query patterns, named TWIX, which imposed minimal restrictions on the user and causes substantial reduction of the search space through a distributed binary labeling technique. The al- gorithm incorporates a holistic ranking scheme of struc- ture and content, named TRANK, to rank and report the top-k results. Furthermore, TWIX benefits from an in- teractive graphical user interface twig query matching. Experimental results on real datasets depict the ranking semantics and efficient filtration of the search space.