Data Stream Management Systems for Computational Finance

MSR-TR-2010-130 |

Computational finance leverages computer technologies to build models from large amounts of data to extract insight. With the increasing amounts of data available to build and refine models, the difference between seizing an opportunity and missing one is the latency involved in processing this data. This article motivates the need for a specialized platform – called a Data Stream Management System (DSMS) – to perform complex processing with minimal latency over large volumes of temporal data. Using a running example from the financial domain, it describes the salient features of these platforms and the advantages they offer.