Friday, November 28, 2008

Complex Event Processing


Carole-Ann Berlioz-Matignon started another small fire storm on her blog about Complex Event Processing.  In running the rabbit trails of I did find another blog on CEP by Tim Bass of Tibco that seems to take a more academic approach to CEP that uses lots of pictures, paragraphs of polysyllabic words and phrases that take some time to read.  Charles Young, he of UK fame, even chimed in with his usual lucid and erudite thoughts.  I did read through the blogs and now I'm pretty sure that no one has a really good grasp on a definitive answer to, "What is CEP?"  

I think my FIRST comment is that neither CEP nor Rulebased Systems nor Neural Nets are "Business" only pursuits, although Fair Isaac, ILOG, Tibco, IBM, Microsoft and others are desperately trying to force everything computer-related into their own pre-defined concept of how that particular tool might be effected in the business world while relegating the rest of the world (physics, chemistry, geology, psychology, medicine, electronics, forecasting, analytics, etc, etc.) to non-essential importance.  They will deny it, of course, but, nevertheless, their actions cry much louder than their words.  If it cannot be forced into their world of business applications then it does not deserve consideration.

Back to the matter at hand:  Wikipedia defines a CEP as "... primarily an event processing concept that deal with the task of processing multiple events with the goal of identifying the meaningful events within within the event cloud.   CEP employs techniques such as detection of complex patterns of many events, event correlation and abstraction, event hierarchies, and relationships between events such as causality, membership, and timing, and event-driven processes."  By it's very name, it is an Event - not something static - and it is Complex in that there MUST be something that about the event(s) that makes it not easily solvable or analyzed.

Since the definition seems to be getting in the way, let's say what it is NOT.  It is not:
  • Simple
  • Time comparisons (only)
  • Rules (only)
  • Processes (only)
  • Neural Net (only)
  • Necessarily Distributed
So, now that we (OK, just me) have said what it is NOT maybe we can say what it might be.  (This part is what is open for discussion.)  How about this?  "A Complex Event is any event in time that is composed of multiple facts - whether static or dynamic - or events the outcome of which defy ordinary logic and the outcome is solvable according to either rules or neural networks and the same set of facts and events always have the same resultant process."  Or something like that.   What a CEP is NOT is BRMS, an inference process (only) nor any one set of things that would constrain the solution.

For example, forecasting is an extremely complex process wherein only the short term can be predicted with any degree of accuracy.  If any financial forecasting package could predict the future within 5% within the next 12 months then the owner would make a small fortune.  You just can't know everything about everything.  Personally, my goal is that between now and October Rules Fest for 2009 I will have made some kind of significant progress on a commercially viable forecasting package.  If nothing else, my goal of understanding exactly HOW to do event-driven-rule-constrained-forecasting will be better.  


Thursday, November 27, 2008

ORF 2009 Blog

ORF 2009 has it's own blog now at - please pay a visit because that's where we will be organizing everything for the next October Rules Fest.  Thanks,


Friday, November 21, 2008

Drools 5.0 M3 is Available

Drools 5.0 M3 is out and they are getting close to a general release.  But there is a TON of work to be done on the documentation that is out of synch with anything including M2 and probably M1, especially on the BRMS.  On the other hand, they are spreading out so five different fronts at one time - this is a MASSIVE undertaking to do with five full-time people and a handful of part-time contributors.  Check out to get the download, documentation (such as it is) and check out what is going to be the future of share-ware BRMS.  (Note:  BRMS by the traditional definition is NOT the same as defined by Drools.)  But, overall, they are closer and closer to what Fair Isaac and ILOG are putting out.  If you rank Fair Isaac Blaze Advisor and ILOG JRules as about a 8, 9 or 10, then Drools would be about a 5 with a bullet.  At their present rate of development, they will catch them (but only if the Big Two stand still) in a couple of years.


Modern Times


Taking time out from product evaluations just to go into idle gear and read some stuff that's been stacking up since last spring sometime.  So, here we go:

Self-Healing Carbon Fiber applications:  Did you happen to see the latest movie on Batman -  Batman Begins?  Did you wonder where Hollywood gets those wacky ideas?  Some are not so wacky.  It seems that that fiber already exists.  There's an article about it in Spectrum, (Oct 2008, p.28) the IEEE magazine that all members receive, called, "Self-Healing Hulls" and it's mostly about self-healing boat/auto/airplane bodies made from carbon fiber.  But, from the description it sounds exactly how Batman's wings operated when they "snapped" in to place.  Wow!

Multi-Core Processors:  Same magazine, page 15.  SAVE YOUR MONEY!!  A report by Sandia Labs says that a 4-core CPU will deliver almost as much speed as an 8-core CPU.  But both deliver more than a 2-, 16-, 32- or 64-CPU machine.  (They didn't say HOW they tested, but, hey!  It's the IEEE combined with Sandia Labs - yes, those labs of Jess fame - so they have to be correct.  Right?)  A 2-CPU machine is on a par with a 16-CPU machine.  The fastest was the 8-core but not by much over the 4-core.  Now, remember, they are checking the speed on only ONE application, not lots of applications nor those requiring lots of threads such as Parallel Rulebased Systems.  But for most rulebased systems, a single application should run fine on 4- or 8-core machines.  If you have 64 or 128 cores, have the other cores working on something else, like a database access or screen updates (which really should have their own processors and memory) or doing complicated math routines.  Sandia Labs, in conjunction with Oak Ridge Labs in Tennessee, are experimenting with stacking memory directly on top of the CPU.

Toshiba Quad-Core Laptop:  Yes, you read that correctly.  QUAD-core on a Laptop.  Called the Qosmio (Kos'-mi-o) it weighs in at 10.8 pounds (about 5Kg) it's a monster.  But, on the up side it comes WITH
  • 4GB RAM
  • 2x250GB HDD
  • 1066MHz Front Side Bus
  • 512MB NVidia Graphics 9600 Video
  • 18.4" WS, 1680x945 Display
  • Built-in Camera with Microphone
  • Blue Tooth and Wireless Built-in
  • $1549.99 USD
OK, it's heavy, it's more expensive than most laptops today but cheaper than my 17" MacBook Pro, Core-2 Duo Apple.  Unfortunately, it also comes with Microsoft Windows Vitri 64-bit OS.  But, guys! - it has FOUR CORES on a LAPTOP!!  See Toshiba for more info.  

Thanksgiving USA:  Finally, Thanksgiving in the USA is next Thursday.  Students and teachers in the USA take the entire week off.  Factories shut down Wednesday noon, sometimes the whole week.  Department stores get ready for "Black Friday" by shutting down on Thursday, stocking up and having sales that start as early as midnight Thursday but usually at 6:00 a.m. on Friday morning.  And the lines begin forming right after midnight but are already formed by 5:00 a.m.  

The question is, who are YOU going to thank for what you have?  Even if you are unemployed, you have access to an internet connection or you wouldn't be reading this.  But, the question remains:  Who do you thank?  Yourself?  Now, that's being silly.  Mom and Dad for making you what you are?  Good start.  The government?  No, WAY, Jack!!  Go back to the original Thanksgiving day and see what they did and why on Wikipedia entry on Thanksgiving.  While you're there, look up Squanto, the Pilgrims and their reasons for forming, migration to Amsterdam, their eventual voyage to what would become New England, the Mayflower Compact, their incredibly poor planning that led them to become dependent on Squanto, etc, etc.  All very interesting...  


Thursday, November 20, 2008

Codeless Programming


InfoWorld (Tom Kandshige, author, and Doug Dineley, editor) ran an article today (20 Nov 2008) on codeless programming, mostly around Coghead and Caspio Bridge (for MS guys).  But they did mention just a bit about the BRMS world on page two.  OK, ok...   It was only one paragraph on page two that nobody else probably read but at least they mentioned KBSC.  :-)  We may do more on this later but for now I'm just inviting comments about the so-called "Codeless Programming" world.  Thanks,


Senior BRMS Consultants / Architects and Head Hunters


OK, I called them by the more socially acceptable name of "Head Hunters" rather than their normal name of "pimp" which would make us, the technical personnel, the "ladies of the evening."  But here's the problem:  These guys have absolutely NO IDEA of what is the difference between someone who went to a one-week school and spent a couple of months on a rulebased (BRMS) project and someone who has 10 - 20 years experience in the field.  None.  Nada.  Zip.  Zero.  To them, a body is a body is a body and if the person proclaims that they hung the moon and stars in the heaven above they they forward the resume on to the next level if that person answers a simple question correctly:  "What is your billing rate and is that all-inclusive?"

In other words, you won't make it to the evaluation round of hiring and working on a project unless you are willing to work for minimum wage.  (Minimum wage in our industry is about $40 or $50 per hour.)  Traveling expenses, usually around $25 to $35 per hour, take up most of that.  So you have to pay your bills with $15 to $25 per hour.  

And the worst offenders seem to be those working with the larger companies such as Sogeti and Accenture.  These companies assign some flunkie from off-shore to do the initial screening for $5 per hour or (usually) so much per resume processed.  This flunkie knows NOTHING about programming nor AI - the flunkie knows only that so many resumes have to be processed every day and if the flunkie happens to find a nugget in the bottom of the pan, so much the better. 

And the part that REALLY gets to me is when they ask, "Well, how well do you know this product?"  This interview had to be scripted by a complete moron!  More than likely they would ask Dr. Forgy if he knew anything about the Rete Algorithm - but that would assume that they had the intelligence to actually know about the Rete Algorithm in the first place.  Any one looking for a job who has been out of work for a few months would self-rate themselves as 9/10 on anything for which the pimp-company is searching.  And the flunkie!  All thy ask is whether you know the name of the vendor / tool and what is your billing rate.

New tactic:  I don't respond to anyone who calls and asks such a stupid question.  If they call and immediately ask my billing rate, I just tell them to go READ THE FABULOUS RESUME!  At least take the time to read it.  I have been in that position when looking for rule team members and I NEVER called or responded to anyone unless I had taken the time to actually read through the resume first.  And I always asked for references AND I always called (or emailed) their references to be sure that we we on the same page.


Sunday, November 16, 2008

ORF 2009 - Call for White Papers


We're starting early this year in our call for papers.  For ORF 2009, regardless of the location and timing, because of the demand from the final session on ORF 2008, we think that it would be best to have ALL of the papers (except for "possibly" a tutorial session prior to ORF 2009) be on the APPLICATION OF RULEBASED SYSTEMS IN BUSINESS AND INDUSTRY.  

Here is the difference between ORF 2008 and anything that BR Forum is doing:  This entire conference will focus on the following:
  • Real World Business or Industry Application
  1. Supply Chain Management
  2. Forecasting and Quantitative Analysis
  3. Insurance
  4. Banking and Mortages
  5. Processing Plant Operations
  6. Geology
  7. Medicine
  8. Science (Chemistry, Physics, Math, etc)
  9. Engineering (EE, ME, CE, etc.)
  • What was needed
  • What was the approach
  • What was the architecture
  • Was the problem solved completely
  • Show the architecture of the rules
  • Examples of rules (yes, screen shots are permissible)
  • Neural Network Solutions to problems are encouraged
  • Solutions should have contained at least 200 rules but there is no maximum
While the main part or most of the rules "might" be business confidential and not displayed, that's OK so long as the attendees are shown the rules that had the most effect on the outcome.   What we do NOT want are the following:
  • NO Product Demonstrations per se
  • NO Toy example systems
  • And nothing to do with a Pet Store UNLESS it was a real Pet Store!  :-)
So, any comments?  This was what most of the attendees at ORF 2008 seemed to want.  I'll probably put out an email later pointing everyone to this blog since most are not using the RSS feed to keep themselves updated.  (We don't have an ORF blog right now but maybe next month.)


Saturday, November 15, 2008

Yaakov and the Sabbath


Yaakov Kohen, my West Texas Jewish Cowboy friend who, at times, serves as my alter ego, has posted the first in what might be a series of blogs on the Sabbath.  He's kind of strange but a lovable kind of guy - typical Texan who chooses to walk his own path rather than "fit in" with the rest of the group.  Anyway, check out his blog when you get a chance.


Tuesday, November 4, 2008

ORF 2009 - Le Deuxieme


ORF 2008

Overall, the conference was quite good - I really enjoyed it and just want to thank ALL of you for turning out. If you had to miss one this year, this was a good one and not one to have missed. The price was right and the place was right and the speakers were, for the most part, right on. :-)

ORF 2009:
 Just some random thoughts - none of which have been discussed with the other guys associated with ORF - but things that I'm just "thinking" about for now. So, anything that I say here is open for discussion, additions, or deletions.

Dates: Tentatively scheduled for October 21st, 22nd and 23rd of 2009.

Location tentatively scheduled for either Fort Worth (TX), San Antonio (TX), Corpus Christi (TX), Miami (FL), San Francisco (CA), Delmar (CA), Nice (France), Paris (France), London (UK), or Munich (Deutchland). Anywhere that is warm to avoid a 22% convention tax PLUS another 8.5% sales tax on everything that we do. Europe would probably mean an October 7th, 8th and 9th.


We'll begin to get abstracts right after the first of the year. Upon the review of the abstracts by the Peer Review Committee abstracts will be temporarily approved. This year we hope to get more technical presentations of "how to" rather than theory. Also, only one Keynote speaker and that one on the first day only.

Full White Paper in PDF format suitable for publication must be submitted by April 15th.

Slides in PDF Format must be submitted by June 15th.

Final Approval to be given to all speakers not later than August 1st.

Diamond (3 max)
Platinum (3 max)
Gold (5 max)
Silver (10 Max)
Bronze (no limit)

Returning sponsors would have first right of refusal on sponsorships at their previous level.

Possibly we could have "Lunch Sponsors" who would, in addition to their regular sponsorship of Diamond or Platinum, want to sponsor Lunch for all attendees AND the Lunch Sponsor would have that time to present a demo of their product. ONLY during that time!!

We could also have "Pub Sponsors" who would be able, in addition to their regular sponsorship of any level, use a main room for drinks, appetizers and spirited conversation.

OK - that's it for now. Let me know what you think and we'll keep everyone in the loop from this blog. Stay tuned.