Bayesian-belief Networks for Supporting Decision-making of the Opening Data by the Customs

Abstract

Open government data initiatives are part of the endeavor process of governments to show that they are accountable and transparent organizations. Opening more datasets to external data analytics providers or other government organizations holds the potential to help governments to improve their processes by promoting a better understanding and enhancing the decision-making. Nevertheless, the decision-making to disclose datasets is challenging. Decision-makers often refuse to open their datasets due to several potential risks. In situations like the Dutch Customs, a dataset can contain competitive sensitive data, and multiple parties have to agree to open it. Given this complex situation, in this paper, we test a Bayesian-belief Network method for supporting the decision to open data. Our work contributes to Customs in their efforts to disclose more datasets and helping decision-makers in the process of evaluating data and defining strategies of how to move from closed to open decisions.