There is a close relationship between formal language theory and data compression. Since 1990's various types of grammar-based text compression algorithms have been introduced. Given an input string, a grammar-based text compression algorithm constructs a context-free grammar that only generates the string. An interesting and challenging problem is pattern matching on context-free grammars P of size m and T of size n, which are the descriptions of pattern string P of length M and text string T of length N, respectively. The goal is to solve the problem in time proportional only to m and n, not to M nor N. Kieffer et al. introduced a very practical grammar-based compression method called multilevel pattern matching code (MPM code). In this paper, we propose an efficient pattern matching algorithm which, given two MPM grammars P and T, performs in O(mn 2) time with O(mn) space. Our algorithm outperforms the previous best one by Miyazaki et al. which requires O(m 2n 2) time and O(mn) space.
|Number of pages||12|
|Journal||Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|
|Publication status||Published - 2004 Dec 1|
ASJC Scopus subject areas
- Theoretical Computer Science
- Computer Science(all)