CS 68191 Masters Seminar / CS 89191 Doctoral Seminar
Spring 2008
Efficiently Querying Reachability on Large Directed Graphs
Ning Ruan
(Doctoral Student Presentation)
Efficiently managing and answering queries on very large graphs is
becoming an increasingly important research topic driven by many
emerging real world applications, range from XML databases, GIS, web
mining, social network analysis, ontologies, and bioinformatics,
etc. Among them, graph reachability query has attracted a lot of
research attention. Reachability query is not only widely used in
graph databases, but also serves as a fundamental operator for many
other graph queries. The major ideas to handle reachability query is
to build indexes based on the reachability labeling. Simply speaking,
each vertex in the graph is assign with certain labels and the
reachability between any two vertices can be quickly determined based
on their labels. Several approaches have been proposed for this
purpose, such as interval labeling (tree cover) and 2-hop labeling,
etc. However, due to the huge number of vertices in many real world
graphs (some large graphs easily contains millions of vertices), the
computational cost and/or the labeling (index) size are still too
expensive for the existing methods to be practically usable. In our
work, we introduce a novel graph structure, referred to as path-tree,
to help labeling very large graphs. The path-tree cover is a spanning
subgraph of G in a tree shape. We demonstrate both analytically and
empirically that the effectiveness of our new approaches.