A word equation in n variables x1,…, xn over an alphabet C is a pair E = (φ(x1,…,xn),Ψ(x1,…,xn)) of words over the alphabet C ∪ {x1,…, xn}. A solution of E is any n-tuple (X1,…, Xn) of words over C such that φ(X1,…,Xn)=Ψ(X1,…,Xn). The existence of a solution for any given equation E is decidable, as shown by Yu. I. Khmelevskiǐ [3] for up to four variables and by G. S. Makanin [6] for any number of variables. However, as shown by A. Kow and L. Pacholskl [4], these impressive decidability results can unfortunately not be matched by efficient algorithms of resolution, except for some restricted classes of equations. In this vein, W. Charatonik and L. Pacholski [1] give a polynomial algorithm, in terms of the equation length |E| - |φ| + |Ψ| for the equations in two variables and very roughly estimate at O(|E|5) the time complexity for solving those in one variable. For the latter, using rather fine combinatorial methods, we give an O(|E| log |E|) algorithm, the best one so far known.
CITATION STYLE
Eyono Obono, S., Goralcik, P., & Maksimenko, M. (1994). Efficient solving of the word equations in one variable. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 841 LNCS, pp. 336–341). Springer Verlag. https://doi.org/10.1007/3-540-58338-6_80
Mendeley helps you to discover research relevant for your work.