We present a derivation of a parallel version of the Knuth-Morris-Pratt algorithm for finding occurrences of a pattern string in a target string. We show that the failure function, the source of efficiency of the sequential algorithm, is a form of search in an ordered domain. This view enables the generalization of the algorithm both beyond sequential execution and the string data structure. Our derivation systematically uses a divide-and-conquer strategy. The computation tree so generated can be mapped onto time, yielding a naive sequential algorithm, onto a processor tree, yielding a parallel algorithm, or onto a data structure, yielding the failure function.
CITATION STYLE
Srinivas, Y. V. (1993). Derivation of a parallel matching algorithm. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 669 LNCS, pp. 323–343). Springer Verlag. https://doi.org/10.1007/3-540-56625-2_21
Mendeley helps you to discover research relevant for your work.