Experience-based early warning of riots. Final report on decision support for international operations

This report is a summary of a project initiated at FOI, aiming at an increased ability to predict threats caused by collective human behaviour (e.g. riots). Predictions are focused upon early identification and interpretation of those signs, signals, behaviours etc. that precede an event or a chain of events caused by collective actions. The project has been studying various ways to find and categorize threat indicator signs, as well as suggesting methodology and computer supported methods to be used mainly by military intelligence units in the field, on international peace-keeping or peace-enforcement type missions. The basic idea is to collect and reuse relevant experiences from earlier missions, and increase the know-how about warning indicators and relevance of events to course of events. The method used is based on so called Case Based Reasoning, building upon analogies for recognition of various alternatives of action. The principle is to store field experiences from preludes of actual riots and other such events menacing to a deployment force. When faced with a new situation, it is hoped that the decision maker, by retrieving and studying similar events from the past, will be helped in assessing serious threats, thereby buying time for preemptive actions.