Thursday, December 11, 2008

Ingomar's Comment

Following is the comment by Ingomar Otter =

Here are my 0.02EUR about how my ORF2009 dream looks like:

Duration: Three days for the core conference. Maybe some extra days if someone comes up with something like the "Bootcamp" the Drools people had this year.
But not as a part of the core conference. If it's "next to a weekend" (starting Monday or ending Friday) would help me personally with the travel.
Maybe if it does not end on a friday we may see less people running away at noon. (just a guess)

Location: No real preference, US/Dallas is fine. Please make sure it's somewhere warm (also on the inside)

* "How to do it" sounds like a good concept to me. Please make this shorter (max 60 minutes, preferably 45) and allow more time for questions and casual exchange.
* I really think ORF2009 should try the OpenSpace format. We have been using this for a few years now and I think it's perfect for the type of exchange I found most valuable at ORF2008.
In a nutshell it turns a conference into a never ending coffee break.
Coffe break = the most valuable moments in a conference when you (by accident) found a peer that shares your intrest/problems etc. It jsut removes the "accident" :-)
See for some more details Anatomy of an Open Space event
So I propose a mixed mode of traditional and OpenSpace sessions, with OpenSpace in the afternoon. Depending of where ORF2009 takes place we could help with organisation/facilitators (we have been doing this on an "ORF" like scale)
* If there are intresting (!) demos/tutorials by vendors then let's have them. Keep them short and sweet so that the presenters have to make a choice what they _really_ want to tell us.
* The Ontology bit was quite intresting to me, I would like to see more of this. I also would like to see more out of the "real" reasearch area and I would not lomit this to "stricly rules".
Everything that makes sense in application areas where rule based systems are applied is welcome.

* For the coffee price we ORF2008 was charged, I will personally prepare it for you ;-)
* We need the cane!
* I was and still am looking forward to the pub nights. It's ORF! :-)
What would help is to have some limited suggestions on the agenda where to go so that it does not become to random. What I mean: "Pub night at X which is located at Y".

ingomar dot otter at valtech dot de

December 2, 2008 9:33 AM

Sunday, December 7, 2008

Call for Papers for ORF 2009


Just in time for Christmas / Channuchah (or Hannucah if you prefer).  Here's the schedule for October Rules Fest 2009:
  • April 15 : Deadline for Abstracts
  • May 1: Response by selection committee
  • July 15: Deadline for White Papers
  • Aug 1: Response by selection committee
  • September 30: Deadline for Overhead (PPT or PDF)
  • October ??: Conference Tutorials and Boot Camps
  • October ??: Conference Begins at 9:00
  • October ??: Conference Ends at 5:30
OK, that sets up a seemingly workable time table.  Any suggestions would be appreciated.   If you are set up to blog here, then please respond with a blog rather than a comment.  Thanks.  Now, what about the talks themselves?  What about the Tutorials?  What about the Boot Camps?  

Tutorials:  I'm suggesting that the Tutorials be on Monday only, all day, 9 - 12 and 2 - 6.  We should be able to teach something about rulebased systems in 7 hours that regular Java, C++, C#, COBOL programmers can take away and build upon for the conference or for the next year(s) study programs.  Anyone wanting to teach or lead a tutorial, PLEASE let me know.  After all, you don't want me teaching everything do you?  :-)  

Boot Camp(s):  This is could be for Monday and Tuesday or Tuesday only.  What we need are vendors who want to have a BC for their product or, better yet, a neutral ground BC (non-vendor-specific) for those who just want a follow-up from Monday.  If the vendors want to have BC on Monday or Monday/Tuesday then I propose that each vendor do whatever they would like to do with a four-hour session; whether a lab, whether demo+lab, whether all demo-instructor-led demo/lab, or just a real BC where the student is expected to already HAVE the software loaded on their laptops and, possibly, already have the labs exercises downloaded from the internet and come to work on solutions.  The possibilities are limitless.  

Or, we could do both on Monday and Tuesday and have two tracks.  ???  One on tutorials to teach rules to experienced programmers who have never done rules and another one that is a boot camp for experienced rulebased KE or RE guys.  

Presentations:  This is the core of ORF!  Those with approved Abstracts will then present White Papers (real ones, not just copies of a PPT) for approval that the attendee could follow should they come into the session late or even miss the session.   White Papers can be rejected for being too simple, too commercial, too abstract (probably not) or even too complex (hardly likely).  Our biggest problem with 2008 was that a few of the presentations were either too simple (managerial level) or too commercial (appeals to the product itself). 

What we want for 2009 is a presentation of  technical thoughts, not just a slide show of product benefits.  Start with a white paper abstract, send it in and immediately begin the white paper.  The white paper should be at least three pages and not more than 20 pages, complete with notes, instructions, forumlae, foundation theory, etc., etc.  In other word, something that would be presentable at any IEEE-AI, AAAI or DSI conference.  After all, this is supposed to be a "technical" conference, not a product presentation for managers or CxO guys (excepting the CTO guy, or course)

Here is a thought on presentations schedule:
  • Wednesday - How to gather requirements.  How to design the architecture for a rulebase.  Knowledge Acquisition.  First steps for a project.  Project design.
  • Thursday - Writing the rules for a particular problem.  Case studies in detail (without revealing any of the company confidential informaiton) would be good.  
  • Friday - Pure Technology Day (PTD): This would be the day to discuss three, four or five of the leading technology problems, proposed solutions, new technology for the next year, where we are going with what we're doing, AND, lastly, again, a conference wrap-up (2008 had a great wrap-up) and Friday supper somewhere that there isn't so much noise that you can't hear yourself think.  - maybe a reservation somewhere? - then off to the Final Pub Night and fly home Saturday morning.
OK, guys - this is YOUR conference and YOU get to determine what we're going to do.  Personally, I would like three days of technology.  But the consensus in 2008 was that we needed more on "How To" lessons that high-level, managerial talks or academic talks.  Whatever...  :-)

BTW, the date and location has not been fixed yet.  We could always move it forward or backward a week if there is a good reason to do so.  Some have expressed a preference not to interfere with Halloween (I didn't know that Halloween was a national holiday but I could be wrong about that as well) which is the reasoning for the earlier week selection.  However, if the majority vote for the following week, so be it.


Thursday, December 4, 2008

In Answer to Paul Vincent


My heartfelt thanks for joining in.  First, let's make it clear that we are NOT a standards committee like OMG and RuleML.  We are NOT a conference aimed at the CxO guys nor the Business Analysts such as BRF.  ORF is a place for those forgotten and often maligned, underpaid and overworked trolls and serfs who work slavishly in the dungeons of corporate empires, keeping things running and helping to improve the bottom line of net-net profit.  This is OUR time and OUR place to come and attend our wounds, apply salve to open cuts and bandage up our fractured egos.  :-)

ORF is THE conference for The Techies.  It is unlike the myriad of business conferences and theoretical/academic type of conferences.  For example, Business Rules Forum (Ron Ross et cie) caters to the CxO guys and the Business Analysts.  He does what, I for one, consider to be a fantastic job.  We are NOT here to compete with that.  (We couldn't if we wanted to!)  Rather, ORF is complementary to BRF in that while BRF is focused on the business problems and showing how they did things from the upper management level, ORF is concerned more with the "how to do it" and "why does it work that way" kinds of problems, things about which the business guys could not care less.

Neither is ORF anything like OMG nor RuleML, the academic branch of this field.  I've been there (at one time I was a member of the OMG PRR group) and we are NOT setting standards.  We probably are very interested in the standards that are being set but we are interested more in HOW to implement those standards in our rulebased systems.

Finally, while we encourage the use of BRMS (Business Rule Management Systems) we are not business only.  ORF is concerned with the "rulebase" approach to solving problems.  We'll even entertain ideas and presentations on neural nets, Prolog, Constraint-Based Programming or anything else that will open up solutions for problems.  We even had one talk that suggested that we not even use an inference engine.  We are looking for another way to think because, ultimately, it is the techies who have to implement what the business guys have paid for.

Something that might not be know is that at ORF 2008 fully 60% of our attendees were NOT engaged in business nor financial systems.  They were doing configuration projects, shipping and transportation, maintenance activities, network planning and configuration, medical diagnostic problems, etc.  Also, 70% or more were from outside of Texas and 25 or so were from outside the USA.

We also received a question from one person who is deeply involved with another conference asking if there was room for another conference on rulebased systems.  My answer is that we'll have to wait and see.  Surely, because one book is written by a world-recognized expert does not mean that someone else should not also write a book on the same subject but with a different viewpoint and audience.  

If ORF continues as it started, it will be because the technical architects and programmers (we used to call them Knowledge Engineers and Rulebase Engineers) really want a place of their own and can justify the minimal expense of attending a conference that is totally dedicated to answering their questions.  After all, ORF costs less than attending a commercial school by any of the many vendors of BRMS.  

So, basically you asked, "What time is it?"  My answer is, "It's time for the Geeks to come out from under the shadow of corporate empires and seek answers and help each other."  ORF 2008 had some of the best speakers in the world; Dr. Charles Forgy, Jason Morris, Gary Riley, Mark Proctor, Dr. Jacob Feldman, Carole Ann Matignon, Carlos Seranno-Morales, and many others who literally have kept this field alive for the past twenty years.  And ORF 2009 will be even better since most of those who spoke and presented this year have already agreed to be there next year.

Rock On!

Wednesday, December 3, 2008

Thoughts on ORF + format suggestion

Firstly, congrats to JCO for the idea of a blog to manage a conference (or probably a better description would be workshop?). For example, academic conferences use to manage paper submissions and peer reviews; commercial ones have read-only web sites or possibly presenter-only sites.

Secondly, I was wondering what the goal / role of ORF was versus the main commercial conference, BRForum (which James T augmented with an EDM Conference) - more business-oriented, and case-studies. The RuleML conference alongside caters for the academic community. So that leaves ORF targeting technical end-users? If so, might be worth stating that in the blog title area (just in case anyone doesn't know!).

One example (IMHO) of a successful technical conference was the original BPMI Think Tank targeted at BPM practitioners, engine vendors, etc. One idea there, which I have seen since at other workshops, and seemed a pretty useful way to ensure audience participation, was the Round Table. Instead of panels with a mostly captive audience, you have topics set in different tables which have an expert moderator (SME as well as good moderator skills, ideally!), with the goal that everyone on the table contributes to the problem definition and solution statements. Examples could be "Web service attributes for decision services", "Stateful rule service uses", "CEP and EDA differences over EDM and SOA", "Standards use cases, or a waste of time?" ... whatever the audience is interested in / organisers want to push. Thie round table idea may or may not be useful for a technical conference - it depends on whether audience participation and contributions are desirable, and whether such discussions are useful.


Tuesday, December 2, 2008


-- Posted by Carole-Ann Matignon

I find that end users get much benefits out of hands-on sessions. If you can find the right combinaison of theoretical presentations, birds-of-a-feather round tables and hands-on practical sessions, attendees are likely to say they have not wasted their time.

There are different ways of doing it.

You can fake it with "how-to" sessions where skilled presenters show how to create a project step by step. This is valuable and the least time consuming. In some cases, you can make it fun by throwing requirements live and see the presenter's approach to design and implement such change.

In some shows like Business Rules Rules, real hands-on sessions are offered (fun labs). This is a mini-training session condensed in 2 hours. The benefit for the user is that they can touch the product and get a real feel for usability and out-of-the-box capabilities. The problem is that you only get to scratch the surface in that amount of time. It cannot provide a broad and deep overview of the product.

I would be curious as to what the expectations are for prospective attendees:
  • Are you looking for a "product evaluation in a box" session?
  • Are hands-on sessions meant to be an illustration of industry best practices?
  • Do you want to see how people use a tool or how you would use that tool?
  • How valuable is it to see how product handle unexpected changes (as they typically happen in real life)?


Thursday, November 27, 2008



Here's what has been suggested by a couple of guys so maybe we should start with considering whether having a three-day conference (like 2008) or a five-day conference that overlaps with Mon-Tu-Wed presentations overlapping on Wed-Thu-Fri and having two tracks.  

One track would be "How to do it" kind of presentations where someone would have 90 minutes to tell, start-to-finish, how a real-life project was actually done.  This would begin with the problem statement, definition compromises, knowledge acquisition, basic design, tools used, some example rules, etc.  However, only one presentation about a bank project, one presentation about an insurance project, one project about a medical diagnosis, one problem about configuration management, one project about ...  You get the idea.  All presentations would be submitted and "The Selection Committee" would pick the ones that we think that most of the attendees would like to see. 

The second track could be tutorials, demonstrations, labs by various vendors, etc.  (I'm not crazy about this one and if nobody wants it we can just not do it.)

The second track could be something practical like different ways of looking at things.  For example, we could do one on ontology (we had two of those at ORF 2008), or maybe another one how to make rules work with Neural Nets or one on using rules with Decision Simulators, etc. 

I would think that Monday morning would begin about 
  • 8:00 - 9:00 registration.
  • 9:00 a.m. to 10:00 Keynote from some industry leader.  ONLY ONE KEYNOTE all week.
  • 10:30 - 12:30 Session 1 for both tracks
  • 12:30 - 2:00 Lunch
  • 2:00 - 3:30 Session 2
  • 3:45 - 4:15 Session 3
  • 4:30 - 6:00 Session 4

  • 9:00 - 10:30 Session 5
  • 10:45 - 12:15 Session 6
  • 12:15 - 2:00 Lunch
  • 2:00 - 3:30  Session 7
  • 3:45 - 4:15 Session 8
  • 4:30 - 6:00 Session 9

  • 9:00 - 10:30 Session 10
  • 10:45 - 12:15 Session 11
  • 12:15 - 2:00  Lunch
  • 2:00 - 3:30 Session 1
  • 3:45 - 4:15  Session 2
  • 4:30 - 6:00 Session 3

And so on to Friday afternoon at 5:00 or so.  Probably (maybe) sessions 8, 9 & 10 would/might be sparsely attended on Friday.  They were WELL attended on Friday for ORF 2008 - probably because we kept the Big Gun until the last session.  :-)  But maybe we could drop out some sessions (9 & 10) and do a wrap-up right before or after Lunch on Friday.  Even with the dropouts, we would have 16 sessions, 8 per track.  Depends on how many really good talks we get scheduled.

Well, there you have it.  The first agendum on the table is the Agenda schedule for the Conference.  :-)


ORF 2009


This blog is set up so that those who are helping to organize the October Rules Fest for 2009 will have place to insert notices and the general public will have the opportunity to comment on the blogs by the organizers. 

For now, the bloggers are limited to those actively participating in helping to make ORF 2009 a reality.  Comments "can" be anonymous, but I wish that anyone making a comment would have the courtesy to sign their name, but they will be moderated so that we don't get junk in the comment pages.  Also, if you wish, you might want to set up an RSS feed to your email so that you automatically get blogs and/or comments.  Enjoy,