Pattern matching algorithm using a succinct data structure for tree-structured patterns

Abstract

Two things are important in developing a fast, memory-efficient graph mining method that extracts characteristic graph structures from Web pages and other tree-structured data. One is tree patterns that can express the features of a graph structure and the other is data structures for tree patterns and for representing tree-structured data. In this paper, we first apply a depth-first unary degree sequence (DFUDS), which is one succinct data structure for an ordered tree, as a succinct data structure for tree patterns that express the features of a graph structure. We then propose a pattern matching algorithm that uses the DFUDS succinct data structure, to determine whether or not a given tree-structured data has features of tree pattern. We also implement the proposed algorithm on a computer and evaluate the algorithm by experiment. The results are reported and discussed. © 2012 Springer Science+Business Media, LLC.