Tim Bass
Thu, 16 Aug 2007 23:41:26 +0000
There are a number of folks interested in CEP and many of them unsure where to get started.* We discussed this recently in the
Yahoo! CEP-Interest eGroup before we got sidetracked with the standard data stream processing obstructionism by our colleagues at
Aleri Labs, who continue to try to redefine CEP in their view of data stream processing. * The debates with those guys seem to never end!
One of the first things people can do to help the CEP community is to share their successful CEP use cases.* Think in terms of difficult distributed network computing problems that are important to your business or organization.* Focus on business solutions, but do not ignore scientific applications of CEP.
Finally, recall that
Dr. Luckham’s et al (and others) original work on CEP was based on solving complex situation detection-oriented problems in a complex distributed network computing environment where the uncertainty and “noise” was much higher than “the signal”.** In other words, focus on problems where the signal-to-noise ratio is very low.
In non technical terms this means focus on use cases where it is both difficult* and very critical to businesses to detect opportunities and threats in distributed networks (like e-commerce, supply chain, sensor nets).
The more use cases and successful application of CEP to real world problems, the better.
Source...