Multivariate Data Generation for Customs Risk Evaluation Tool

Today, vast volumes of goods are transported all over the world in containers. Customs authorities are charged with detecting smuggling and can find indications of this by screening documentation on containers that is by law provided by shippers and carriers. In the Contain project, Decision Support Systems for customs to do this risk profiling are developed. In order to test these systems, we have developed a tool (ENS-simulator) to provide simulated input to the profiling tool. In this paper, we present the simulation tool and describe the method used for generating a high rate of messages in a realistic way, to represent a typical message inflow at a large customs risk assessment center.