Our paper “How not to drown in a sea of information: An event recognition approach” was accepted at IEEE Big Data 2015
Elias Alevizos, Alexander Artikis, Kostas Patroumpas, Marios Vodas, Yannis Theodoridis and Nikos Pelekis
Maritime monitoring is a typical Big Data problem where hundreds of thousands of vessels across the globe transmit messages about their location, speed and other information. We have developed a system for online vessel tracking that performs, as a first step, a high-rate but accurate trajectory compression. Subsequently, the compressed trajectories are analyzed by a complex event recognition engine, promptly reporting alerts to maritime authorities. To deal with realistic maritime event patterns, we seamlessly integrated spatial andtemporal reasoning for online event recognition. The system is evaluated on real data from the Greek seas.