A Stochastic Model for Document Processing Systems
CAMPUS IAI, B.P. 2263 Libreville, Gabon, Pierre Moukeli Mbindzoukou, Jocelyn Nembé · International Journal of Information Engineering and Electronic Business · 2016
This work is focused on the stationary behavior of a document processing system.This problem can be handled using workflow models; knowing that the techniques used in workflow modeling heavily rely on constrained Petri nets.When using a document processing system, one wishes to know how the system behaves when a new document enters in order to give precise support to the manager's decision.This requires a good analysis of the system's performances.But according to many authors, stochastic models, specifically waiting lines should be used instead of Petri nets at a strategic level in order to lead such analysis.The need to study a new model comes from the fact that we wish to provide tools for a decision maker to lead accurate performance analysis in a document processing system.In this paper, amodel for document management systems in an organization is studied.The model has a static and a dynamic component.The static one is a graph which represents transitions between processing units.The dynamic component is composed of a Markov processes and a network of queues which model the set of waiting-lines at each processing unit.Key performance indicators are defined and studied point-wise and on the average.Formulas are given for some example models.