Discovering patterns in flows: A privacy preserving approach with the ACSM prototype

Abstract

In this demonstration, we aim to present the ACSM prototype that deals with the discovery of frequent patterns in the context of flow management problems. One important issue while working on such problems is to ensure the preservation of private data collected from the users. The approach presented here is based on the representation of flows in the form of probabilistic automata. Resorting to efficient algebraic techniques, the ACSM prototype is able to discover from those automata sequential patterns under constraints. Contrary to standard sequential pattern techniques that may be applied in such contexts, our prototype makes no use of individuals data. © 2009 Springer Berlin Heidelberg.