The RAHS system seeks to give early warning of possible emerging strategic-level surprises, typically on the 2-5 year time horizon, so that policy choices can be made, and preemptive and preventive steps taken where necessary. The system is not a tool to monitor or direct on-going crises.

It is not a tool for prediction. It cannot foretell discrete events. It is designed more as a tool for anticipatory analysis, to pick up weak signals or outliers pointing to worrying trends, crises and major turning points. The tool cannot replace the human analyst. It is meant to augment the analyst by helping automate some of the more tedious aspects of research work, so that the analyst can focus on the more important analytical tasks. It cannot replace instinct and it cannot ask the right questions.

The RAHS system does not perform data mining. Data mining attempts to extract patterns automatically from databases. The idea here is that the patterns extracted can then be used to identify threats automatically against incoming data. Past efforts around the world have shown that fully automated early warning systems can result in high false alarm rates.

The RAHS system is not an automated early warning system. The philosophy behind RAHS is to augment and enhance the human analyst's ability to see and discover patterns from data. Towards this end, RAHS system development will continue to tap advances not only in information technology but, more importantly, advances in theories of human decision making from diverse fields including cognitive psychology and the social sciences.

>>The wider RAHS programme