摘要：Many fuzzy automaton models have been introduced in the past. Here, we discuss two basic finite fuzzy automaton models, the Mealy and Moore types, for lexical analysis. We show that there is a remarkable difference between the two types. We consider that the latter is a suitable model for implementing lexical analysers. Various properties of fuzzy regular languages are reviewed and studied. A fuzzy lexical analyzer generator (FLEX) is proposed. 1.) The work reported here has been supported by the Natural Sciences and Engineering Research Council of Canada grants OGP0041630 and the Project 11281 of the Academy of Finland.
