Lecture Notes in Computer Science, 1996, Volume 1113/1996, 382-393, DOI: 10.1007/3-540-61550-4_164

Polynomial automaticity, context-free languages, and fixed points of morphisms (Extended abstract)

Ian Glaister and Jeffrey Shallit

View Related Documents

Abstract

If L is a formal language, we define A L (n) to be the number of states in the smallest deterministic finite automaton that accepts a language that agrees with L on all inputs of length le n. This measure is called automaticity. In this paper, we first study the closure properties of the class DPA of languages of deterministic polynomial automaticity, i.e., those languages L for which there exists k such that A L (n)=O(n k ). Next, we discuss similar results for a nondeterministic analogue of automaticity, introducing the classes NPA (languages of nondeterministic polynomial automaticity) and NPLA (languages of nondeterministic poly-log automaticity). We then show how to construct a context-free language with automaticity arbitrarily close to the maximum possible. Finally, we conclude with some remarks about the automaticity of sequences, focusing on fixed points of homomorphisms.
Research supported in part by a grant from NSERC. Please direct all correspondence to this author.

Fulltext Preview

Image of the first page of the fulltext document