A generalisation of Parikh's theorem in formal language theory

24Citations
Citations of this article
1Readers
Mendeley users who have this article in their library.
Get full text

Abstract

We show that when a family of languages F has a few appropriate closure-properties, all languages algebraic over F are still equivalent to languages in F when occurrences of symbols are permuted. At the same time, the methods used imply a new and simple algebraic proof of Parikh's original theorem, directly transforming an arbitrary context-free grammar into a letter-equivalent regular grammar. Further applications are discussed.

Cite

CITATION STYLE

APA

van Leeuwen, J. (1974). A generalisation of Parikh’s theorem in formal language theory. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 14 LNCS, pp. 17–26). Springer Verlag. https://doi.org/10.1007/3-540-06841-4_49

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free