On global maritime oil piracy: an association rules analysis

Abstract

AbstractMaritime piracy is a global problem that disrupts oil flows and products. This research attempts to identify potential association rules between the features of hostile acts against tankers. To that end, the Apriori algorithm is used. We use data from the National Geospatial-Intelligence Agency and modify them in a basket case style. Essential features of an attack are examined, such as the closeness to the coasts, the cruising status of the tankers, and the time neighboring other attacks. Despite the perception we have of them, it turns out that pirates follow conservative strategies. It is found that success is strongly coupled with the safety of the attack by striking near the coast and on anchored tankers. The produced association rules include different time-neighboring conditions and no values regarding the attack month. This implies no strong preference for starting an attack, close or far from a previous episode, and no seasonality.