Incremental Construction of Minimal Acyclic Finite-State Automata
We have tries, which is nice. But we would like better: directly build a minimal acyclic automaton for a given language (not series: we stay in B).
See this paper: http://www.mitpressjournals.org/doi/pdfplus/10.1162/089120100561601
See also if we could have Revuz's algorithm as a specialized version of minimize.