In this paper we describe an event-type extractor on top of a distributed search engine. We apply this event-type extractor in a case study concerned with assisting maritime security operators to assess potential risk factors of ships. Based on a corpus of maritime-related press releases we automatically investigate the history of ships as they enter an area of interest. The performance of the system is evaluated with a task-oriented focus on a set of vessels with known risk factors, and typical behaviour is evaluated by batch-processing a large set of vessels.