CEP, then “BEP”, and now “AEP”
Posted by Paul Vincent
In addition to the de facto term “Complex Event Processing” (CEP) we have had IBM trying to differentiate a subset called “Business Event Processing” when they acquired Aptsoft, and now analyst Charles Brett introduces comments on “Analytical Event Processing“.
But is “Analytical Event Processing” a very useful term? Indeed, what would “non-analytical” event processing be? Guess-based event processing? Random-reaction event processing? Or is this something to do with the fact that some CEP applications (usually regardless of CEP technology - stream processing, rules processing, etc) do what has been described as “event-based analytics“?
Reading Merv Adrian’s discussion with Charles on the topic, there seems to be some thoughts that event processing (whether A-, B-, or CEP) might start to reign in the dominance of data(base)-oriented processing. Already happening, for sure, although as the authors infer, the RDBMS will continue just fine as a data archive, backing store, etc. Reading between the lines, the discussion on the term could also be due to mis-/over-reading some vendors’ attempts at differentiation (and/or said vendors “competitive issues” in the CEP market…).
As to the future of the “AEP”, we shall wait and see! (Or rather, regardless of terminology, CEP vendors will just carry on delivering event processing solutions to customers via a number of different technologies, including predictive analytic technologies where relevant).
Tags: AEP
Analytics, Complex Event Processing (CEP), Trends | 15 November 2009 |
4 Comments
Other Links to this Post
RSS feed for comments on this post. TrackBack URI
By James Taylor, November 17, 2009 @ 13:57
Paul
Why do we need to classify it? Surely “event processing” or “event driven architecture” is both necessary and sufficient? It is a move away from process-centricity and towards event-centricity that is critical to my mind. The decision making involved in deciding how to process an event should be Business-oriented, might be Complex and will increasingly be Analytical. BEP, CEP, AEP - seems like a distinction without a difference to me.
By brenda michelson, November 17, 2009 @ 15:44
Hi Paul,
You’ll enjoy this, Starview, the ‘coiner’ of AEP, has posted on what it is, and why (of course) it’s better than plain ol’ CEP engines: http://starviewtech.wordpress.com/2009/11/16/what-is-analytical-event-processing-aep/
Best,
Brenda
By Paul Vincent, November 17, 2009 @ 16:56
James - the term CEP differentiates continuous event processing over say BPM/workflow event processing. Yes they are all processing events, but are suited to different problems. We are focussing on CEP here…
An analogy might be: why do we differentiate between “analytics -driven decisioning” and “inference based decisioning” - they are both “decision support technologies” - we do ‘cos they have different roles to play…
On the other hand, if you are pointing out that the progression CEP - BEP - AEP must end in EP, perhaps you are right!
Cheers
By Paul Vincent, November 17, 2009 @ 17:13
Hi Brenda - Indeed, enjoyed… hopefully I wasn’t being too rude with the observation “…the term could also be due to mis-/over-reading some vendors’ attempts at differentiation”
But wait, have we not seem some other attempts at jargon for this space? Ah yes - “RTAP” - see http://mervadrian.wordpress.com/2009/05/18/cep-the-tech-that-dare-not-speak-its-name/ !!!
Cheers