A Fast Approximate String Searching Algorithm

Abstract

In both approximate and exact string searching algorithms, the shift distance at the skipping step plays a major role in the performance of string matching algorithms. A new algorithm called the Modified Character-Weight Algorithm (MWA) has been developed to test the effect of the shift distance on the performance of approximate string matching. An experiment was performed comparing the performance of the MWA with that of Mhashi’s Character-Weight Algorithm (WA) using English text of size 1,005,077 characters. Using the average number of comparisons and the clock time as evaluation criteria, the MWA algorithm used only about 4% to 15% as many comparisons as the WA algorithm and about 10% to 35% of the clock time.