Towards the Stream Analysis Model in Grid-based Zero-Latency Data Stream Warehouse.

Abstract. Recent emerging applications increasingly generate continuous, larger amounts of valuable data. The demand of conducting advanced analysis over fast and huge data streams to capture trends, patterns, and exceptions become crucial. However, fully extracting the latent knowledge within the data stream is a challenging task because of insufficient technology. While Data Warehouse (DWH) technologies have resulted in considerable information processing efficiencies, there is still a significant delay in the time to deliver mission critical information to data consumers. Traditional data stream processing focuses on statistical approaches hence produces approximate results. In this paper, we introduce the Stream Analysis Model with a Grid-based Zero-Latency Data Stream Warehouse (GZLDSWH) framework which allows to perform analytical processing on continuous data streams and to trigger relevant actions depending on patterns discovered in event streams without using statistical approximation. Essential data stream elements are captured, analysed on the fly and finally evaluated to detect abnormalities while the entire data streams are stored within a Grid and integrated into a virtual DWH for further analysis in the case of ambiguity or uncertainty. 1