Statistical Data Analysis of Continuous Streams Using Stream DSMS
Several applications involve a transient stream of data which has to be modeled and analyzed continuously. Their continuous arrival in multiple, rapid, time-varying and possibly unpredictable and unbounded way make the analysis difficult and opens fundamentally new research problems. Examples of such data intensive applications include stock market, road traffic analysis, whether forecasting systems etc. In this paper, the authors have used a Data Stream Management System tool-Stanford STREAM to model and analyze data from two different application domains-Road Traffic analysis and Habitat Monitoring analysis. Based on the results, they discuss advantages and disadvantages of STREAM.