To collect and reuse experiencies of early warning of riots. Basis for decision support for international operations

In this we report suggest a working methodology and computer support that may contribute to enhance the ability to predict threats caused by collective human behaviour (e. g. riots). Predictions are focused on early identification and interpretation of such signs, signals and behaviours etc. which precede an event or chain of events caused by collective action. Our target users are mainly a military intelligence units in the field at international missions of the peacekeeping or peace enforcement type. The main idea is to reuse relevant experiences and the aim is to increase the know-how about warning signs and indicators, and their influence on the course of events. The working method originates from so called "Case Based Reasoning" (CBR) which draws upon analogies for recognition of various alternatives of action. The principle is to store field experiences early on from prelude to riot and other for the deployment force menacing events related to collective behaviour. By retrieving when faced with a new situation, similar observations and studying these, the decision maker will hopefully be helped in judging probable and serious threats, thereby buying time for own preemptive actions. This report discusses, from a user´s point of view, both methodologies and computer support methods to collect, retrieve and present experiences in the form of cases consisting of similar events. To be able to design this kind of computer support, we discuss the needs and demands on tools and information structure for both short and long term development.