An Efficient Algorithm for Generating Super Condensed Neighborhoods

Luís M. S. Russo, Arlindo L. Oliveira


Indexing methods for the approximate string matching problem spend a considerable effort generating condensed neighborhoods. Here, we point out that condensed neighborhoods are not a minimal representation of a pattern neighborhood. We show that we can restrict our attention to super condensed neighborhoods which are minimal. We then present an algorithm for generating Super Condensed Neighborhoods. The algorithm runs in $ O(m\lceil m / w \rceil s)$ , where $ m$ is the pattern size, $ s$ is the size of the super condensed neighborhood and $ w$ the size of the processor word. Previous algorithms took $ O(m\lceil m / w
\rceil c)$ time, where $ c$ is the size of the condensed neighborhood. We further improve this algorithm by using Bit-Parallelism and Increased Bit-Parallelism techniques. Our experimental results show that the resulting algorithm is very fast.

Luis Manuel Silveira Russo 2007-12-02