Querying Graph-Structured Data

Abstract

Graphs have great expressive power to describe the complex relationships among data objects, and there are large graph datasets available such as Web data, semi-structured data and XML data. In this paper, we describe our work on querying graph-structured data, including graph labeling methods, reachability joins, and graph pattern matching. We show that we can base on the graph labeling of complex XML and semi-structured data to process path queries and we devise join primitives for matching graph patterns. Novel aspects about using such join primitives for graph pattern matching are addressed.