Data Stream Analysis Current data mining techniques rely on large volumes of stored data from which to draw correlations and create projections. Real-time applications frequently generate more data than is feasible to store. Fortunately, new algorithms for analyzing live data streams are emerging which allow the user to better predict "bursts of activity", detect and track events, and validate data quality. This sophisticated temporal clustering technique is being developed to illuminate trends as they occur, and to project which events are developing.