Using Finite State Automata for Sequence Mining

Abstract

We show how frequently occurring sequential patterns may be found from large datasets by first inducing a finite state automaton model describing the data, and then querying the model.