Discovering event evidence amid massive, dynamic datasets

Automated event extraction remains a very difficult challenge requiring information analysts to manually identify key events of interest within massive, dynamic data. Many techniques for extracting events rely on domain specific natural language processing or information retrieval techniques. As an alternative, this work focuses on detecting events based on identifying event characteristics of interest to an analyst. An evolutionary algorithm is developed as a proof of concept to demonstrate this approach. Initial results indicate that this approach represents a feasible approach to identifying critical event information in a massive data set with no apriori knowledgeof the data set.