Data-driven intelligence analysis

Data-driven analysis is becoming increasingly relevant within the intelligence analysis domain. The amount of available data from sources such as signal intelligence and open source intelligence is growing steadily, which seems to require a shift from theory-driven analysis to a more data-driven analysis. In this report, we shed light on the term data-driven analysis from the perspectives of intelligence and computer science. When using data-driven analysis, the focus can be on either the interpretation of existent data or on the creation of models aiming at making predictions or classifications when new data becomes available. A number of techniques for those purposes are described, divided into the categories of online analytical processing (OLAP), data mining, and knowledge-based systems. A large number of examples on data-driven analysis can be found in open literature, including areas such as terrorism, cyber security, natural disasters, and armed conflicts. A selection of such examples is presented, with the purpose of providing the reader with a better idea of how data-driven analysis can be used for strategic and tactical intelligence analysis.