Benchmarking at ORF 2009

Greetings:

There have been many suggestions as to the problems for the vendors at October Rules Fest this year.  I have submitted one already (Einstein's Puzzle) and the "beginning" of another (Miss Manners 2009) but I have not done the data on the last one.  I really was hoping that an enterprise-size vendor (FICO, ILOG, Haley, Pega, ???) would find someone on staff that is not too busy and task them with this simple but time consuming job.  Didn't happen.

Geoffrey De Smet suggested that we use the benchmark problem from the ITC competition, http://www.cs.qub.ac.uk/itc2007/index.htm that had to do with Exam Registration.  On the benchmarks page they said that it was running between 300 to 500 seconds on a "modern PC using Windows XP."  I would think that 5 - 10 minutes would be a bit long for my tastes but you never can tell.  Personally I prefer old-fashioned Unix (or UNIX if you're that old) rather than Windoze or even Linux.  

I have looked at this particular problem / benchmark and it appears to be more of an Operational Research (OR) problem in optimization than a rulebase problem.  We "could" require that the solution be done in a rulebase but those who have OR engines would scream bloody murder if we did.  

However, I leave it with you in the interim to determine what should be a good rulebase problem and what does not lend itself to simple "data smashing" of a typical "business" problem.  The problem should be fairly complex and require a bit of work on the part of the engine itself rather than depending on the operating system and database to do the work.  Please either comment (or if you have authorship privileges then blog) some kind of response to this.  Whatever is decided should be a problem that is long on complexity and short on blandness.  

SDG

jco

ORF 2009 Abstracts Due in One Week

Greetings:

Short blog:  There is only one more week until the Abstracts are due for October Rules Fest 2009.  We have received about 11 or 12 right now and we have enough to do the conference with some really long session times.  :-)

BUT, we need more quality presentations.  We have five more persons that have promised abstracts by the 30th of April but we need to get those in so that the committee can review them without being too rushed.  It should not take more than five minutes or so to write up the abstract once the thought process of considering what to do and how to do it are completed.

What happens is that we should get about 30 abstracts in, cull out five or so that aren't actually applicable to the conference - although what we have so far is looking great - and then five or so will pull out before the actual white paper is due on July 1st.  After that, if the white paper has been submitted, they speakers usually complete the overheads without much of a problem.  Even if we get 30 or 40 that are good, we will have the time to present them.

So, get busy and get the abstracts submitted ASAP.  Thanks,

SDG

jco

ORF 2009 Now Accepting Registrations

Greetings:

Great News!!  You can now register and get the Early Bird Discount at http://www.OctoberRulesFest.org and click on Registration.  If you are interested in sponsoring, PLEASE fill out the for for Sponsor interest.  

Also note that we are filling up the speaker slots so if you have a talk PLEASE send that in to jco@kbsc.com ASAP.  thanks,

SDG

jco

Registration is NOT Ready Yet

Greetings:

http://www.OctoberRulesFest.org is not ready YET to accept applications but it should be by the end of the week.  I'll post a reminder when it is ready.  Remember, there is a deep discount for Early Bird signups.  One other thing:  If we don't get at least 20 for the tutorials on Monday we won't have that part, just the conference that starts on Tuesday.

Some of the speakers from last year have already signed up for this year and you can see their profile on the "Speakers Profile" tab.  If you want to speak, PLEASE submit an abstract BEFORE the 15th of April.  Thanks,

SDG

jco

Problem Number Two

Greetings:

This one took a bit more thinking.  For the second problem, which we DO require that it be solved with a rulebase, we have an updated Ms. Manners (Miss Manners for this century) in which we have 

12 tables with 12 settings (or seats)

 

144 guests who have 1, 2 or 3 hobbies randomly assigned from a selection of 5

Hobbies are Tennis, Golf, Motorcycles, Chess and Poker

24 are politicians, 

12 Democrats

6 Democrats are female

12 Republicans

4 Republicans are female

24 doctors, 

8 pediatricians

6 Pediatricians are female

8 surgeons

4 Surgeons are female

8 osteopaths

2 Osteopaths are females

24 socialites

18 socialites are female

24 sports stars

8 basketball stars

4 basketball stars are female

12 football stars

4 baseball stars

24 teachers

10 teachers are female

24 Programmers

  8 Mac programmers

3 Mac programmers are female

8 Windows Programmers

6 Windows Programmers are female

8 COBOL Programmers

2 COBOL programmers are female

The objective is to put at each table

1 democrat 

1 republican 

2 doctors at each table but NOT two of the same kind

2 socialites at each table

2 sports stars at each table but NOT two of the same kind

2 teachers at each table

2 programmers at each table but NOT two of the same kind

Meanwhile, maintain a boy-girl-boy-girl seating arrangement

Also, each person MUST have someone (left or right) who has the same hobby.

A log should be kept for each table as it is completed.  We expect that some tables will be filled and then changed as needed.  If no solution exists then the rulebase should report that as well.  Perhaps we need two sets of data:  One that has a solution and one that does not just to see how the programmer and the rulebase deal with failure.

Data for this problem will be furnished on request so that we are all working from the same "possible" solution.  It would be really nice if the participating vendors were to have some nice college intern create the data.  Finally, the solution will be checked for accuracy and completeness.  Actually, it should be a simple enough problem and the rules, while not terribly complex, are left up to your discretion.  Ergo, nothing forced EXCEPT for a real solution.  Remember this is Ms. Manners, not Miss Manners.  Ms. Manners is an OO-Rulebase Problem.

Hopefully, all of the sponsors (vendors) at ORF will have the solutions for these problems ready for the attendees.  :-)

SDG

jco

Abstracts Guidelines

Greetings:

First things first: Abstracts (Tutorials or Conference) are due BEFORE April 15th. Please, please, please don't wait until the very last minute to try and think something up. You wouldn't let your kids (nor your students if you are a teacher) get away with something like that, would you? So, be kind to the Review Committee and submit early. You may submit TWO talks if you like and/or TWO presentations from the same company.

You can submit more than two per company for now and IF we have room later we can add them in - but for now with fewer talks taking more time it is going to be a tight squeeze. Also, if you plan on presenting during the Monday Tutorials, then you can, of course, present another talk during the regular conference.

Following is a format that we would like for you to use.

{title} [Centered and Bold]
[1, 2 or 3 column table containing the following]
{author1}
{company1}
{snailmailaddress1}
{city1, state1, postalCode1, country1}
{email1}
{website1|

{author2}
{company2}
{snailmailaddress2}
{city2, state, postalCode2, country2}
{email2}
{website2}

See http://orf2009.blogspot.com/2009/02/white-paper-or-conference-paper.html for more details.

One or two paragraphs, about 100 - 300 words, that best describe the Who, What, Where, When, Why and How of the White Paper itself. The abstract and references (the ones that you have so far) can be on the same page. References can be either an added page(s) or, if only two or three, on the same page as single-line-spaced footnote.

Use 12 point type and double-line-spacing for the abstract for now. (If in doc or docx format - or any other format that can be imported into MS Word - we can reformat but it would be nice if you did it up front.) 1" margin top and bottom with 0.5" margins on each side. Submit in either doc, docx or pdf format.

White papers will follow a similar format but submission characteristics will change slightly.

The conference papers will be a valid, refereed white paper that academics can use for their publications credentials. The Monday Tutorials have to be approved but are not refereed in the normal sense of the word.

Please feel free to contact info@OctoberRulesFest if you need anything more than what has been blogged already on the subject. So far we have the following abstracts that have been submitted with more promised this coming week:

Jason Morris
"Implementation of a Modified Propose and Revise Problem-Solving Method Using Jess"

Dr. Charles Forgy
"Possible Extensions To Rulebase Languages To Support Cooperating Rules Sets That Run In Parallel" (Or something like that)

James Owen
"Updating Rulebased Forecasting (RBF) To Use Modern Rule Engines"

Dr. Rick Hicks
"Automated Verification Testing for Propositional Logic Systems"

Dr. Jacob Feldman
"Rule Violations and Over-Constrained Problems"

Titles and abstracts may change "slightly" between now and July 1st when the final white papers are due. If you have some of the references ready now, please include them. And hurry! :-)

SDG
jco

ORF 2009 at First Blush

Greetings:

OK, this is a second call for white paper abstract.  It should be about 100 - 200 words that briefly (but completely) tell what will be the contents of your white paper.  This is due NOT LATER THAN 15 APRIL 2009.  So, hurry and get that part to us so that the "committee" can begin their review.  

We have received only one abstract to date but we have promises of about 10 so far.  PLEASE let me know SOON if you plan on speaking.  There were 12 or 13 who signed up to speak but this page did not report who they were - all voting was anonymous.  :-(  Thanks. 

ORF 2009 IS BACK IN DALLAS for the last week of October.  Three and 1/2 days this time, Tuesday morning through Friday noon.  Full details as they are available will be on the web page to go up later this week.  Early sign-up discounts for both the conference and tutorials.  Also, discounts for group memberships.  Probably only one Keynote Speaker this year unless we feel the need for more - let me know what you think.

TUTORIALS will be all day on Monday AND will be available only if we get at least 25 persons to sign up for them.  (The room will be the same room as the conference so we have to have enough to pay for the room, coffee, etc.)  However, we are going to try and persuade the sponsoring vendors do one-hour tutorials (four of them) in the afternoon.  The morning sessions will be introduction to rulebased systems and BRMS taught by several, very senior ORF members.  (That will be on the web page later this week.)

EARLY SIGNUP for the conference will begin later this week.  This will be the cheapest training course for rulebased systems and BRMS in the known world.  Most of the technical speakers from last year will be returning, including Dr. Forgy, Gary Riley, Dr. Feldman, Dr. Hicks, Edson Tirelli,  and Mark Proctor.   Subjects will range from Complex Event Processing (CEP) to RuleBased Forecasting (RBF) to Parallel Rulebased Systems to Dr. Forgy's new and improved algorithm that was announced earlier here on the page.

PUB NIGHT will be a published, established place where we will have reservations in advance.  Last year most of the Q&A was at Pub Night until closing.  What a rush!  But this year we can control the music level so that we can talk without having to shout at each other.

SPONSORS need to sign up soon.  The fees are along the same line as last year and we have sent the fee schedule to most of you already.  If you did NOT get one, let me know.  Without our sponsors, the conference will not happen.

So, come early and stay late.  Most of all, learn something and have fun doing it.

SDG

jco