Peculiarity Rule Mining for Maritime Traffic Outlier Detection

Abstract

The large maritime traffic volume and its implications in economy, environment, safety, and security require an unsupervised system to monitor maritime traffic. In this work, a method is proposed to automatically produce peculiarity maritime traffic representations from historical self-reporting positioning data, more specifically from automatic identification system data. Finding outliers in collection of patterns among AIS trajectories is critical for real time applications ranging from military surveillance to transportation management. So that we proposed the method to measure the similarity between trajectory points and clusters, cluster relative distance and cluster angular distance. Finally, the novel method presents a light and structured representation of the maritime traffic, which sets the foundations to real-time automatic maritime traffic monitoring, anomaly detection, and situation prediction.