Wednesday, February 25, 2009

Einstein Puzzle in Microsoft Solver Foundation

As promised in my last post (http://geekswithblogs.net/cyoung/archive/2009/02/24/129639.aspx), I had a look at the Express edition of Microsoft Solver Framework (http://code.msdn.microsoft.com/solverfoundation), and used it to implement the Einstein puzzle using the built-in CSP solver and C#. Creating the solution proved particularly easy for the simple reason that Microsoft provides a worked example of a politically corrected ‘Zebra’ variant in their documentation (no tobacco or beer in sight). You will find this as Sample 8 in the SFS Programming Primer document that comes with the download.

To read more and see the code, visit http://geekswithblogs.net/cyoung/archive/2009/02/25/129672.aspx

Tuesday, February 24, 2009

Constraint Satisfaction and Rete

A few days ago, Daniel Selman of ILOG (now owned by IBM) published a solution to the Einstein puzzle. See http://blogs.ilog.com/brms/2009/02/16/einsteins-puzzle/. He did this in response to a challenge from James Owen who is one of the organisers of the October Rules Fest conference. James invited the various vendors who have involvement in the conference to provide a solution using whatever approach they deemed best. See http://orf2009.blogspot.com/2009/02/puzzle-1-for-conference.html.

Daniel suggests that, if ILOG had used a Rete rules engine, they would have had to implement 'convoluted rules' to solve the problem. Daniel is correct, but I thought it would be interesting to look a little deeper at why this is the case.

Read more at http://geekswithblogs.net/cyoung/archive/2009/02/24/129639.aspx.

Thursday, February 19, 2009

Job Board

Greetings:

While this is NOT a jobs board, we probably should start one for rulebase consultants.  For now, here is one that I know about:

BIZRULES is looking for a Haley Rules Modeler *

The Haley Rule Modeler designs and implements the Haley Rules using the Haley data model and rules. The Haley Rule Modeler is expected to have extensive experience implementing the Haley rules engine and should have experience implementing the Haley rules engine in the context of Seibel implementations.

* by "Haley Rules" the client means
- RuleBurst Rule Engine (BRE)
- SoftLaw Rule Engine (BRE) / Expert System (ES)
- Haley Office Rules
- Haley Expert Rules
- Haley Business Rule Engine (BRE)

This is for a long term project. RULEBURST/SOFTLAW experts anywhere in the world (Australia, UK, Canada, USA, etc.) are welcome to apply for this challenging opportunity! Relocation available for top candidates.

If you are interested and have experience with RULEBURST / SOFTLAW / HALEY RULES contact us or send your resume to

Careers8@BizRules.com

+1 305.994.9510

Call or email Rolando directly - not to me.

SDG
jco

Wednesday, February 18, 2009

Solutions and Videos

Greetings:

Daniel Selman of ILOG presented the first solution by a vendor for the Einstein's Puzzle.  He has posted it at http://blogs.ilog.com/brms/2009/02/16/einsteins-puzzle/comment-page-1/#comment-1334 and he did it without using JRules.  I had two friends do this in about 20 minutes using only pieces of paper (not very scalable but it's one way) and some others doing it with spreadsheets - not scalable either.   Problem # 2 will be business related and probably be some kind of real-world constraint problem - but that will be next week.  You won't be able to do it with pieces of paper, for sure.

AKSO, Volker Grossman of Innovations Technology (Visual Rules) has put up a UTube video at http://www.youtube.com/watch?v=lfzzHM0uT1o that explains business rules and might be helpful to "noobies" in our industry.  And it's free.  Remember, I said "business rules" and not "rulebased solutions" - something that can be quite involved and complicated.  

But, check them out and see what you think.  Also, check out the other blogs listed on the left side of our page for more discussions on rulebased systems, CBP, CRM, RBF, and lots of other solutions.

The sign-up page for ORF 2009 is coming this week so check by in on Monday and sign up for October Rules Fest 2009.  Bigger and better and still in Dallas.  Next year in Nice.  (maybe?)

SDG
jco

Wednesday, February 11, 2009

Puzzle #1 for Conference

Greetings:

OK - the Einstein Puzzle is NOT a business problem BUT it represents a problem that is best solved by rules and/or constraints.  The complete description can be found  at http://www.stanford.edu/~laurik/fsmbook/examples/Einstein'sPuzzle.html and we will review the various solutions at the conference.  I would have suggested the venerable Monkey and Bananas (MAB) but that one is way too easy and has already been solved many times over.

However, if each sponsor / vendor takes the time to have someone construct an elegant solution in whatever tool is deemed most appropriate, then the attendees can see how a fairly complex problem can be solved using whatever tool is handy.  I'm sure that the times to run such a simple problem will be extremely fast but what we can see will be the elegance (or lack thereof) in each vendor's tool box.  

Comment at will - but for now I haven't heard any suggestions whatsoever from anyone.  :-)  Puzzle # 2 should be a bit more complex and without the guidance of Einstein's Puzzle.  More on that next month.

SDG
jco

Tuesday, February 10, 2009

White Paper or Conference Paper

Greetings:

I would that folks who are "authorized" to blog here would blog rather than just comment.  On the other hand, some comments should be comments.  :-)  Here is the conundrum:  Some want to call what we are presenting at the conference a "Conference Paper" while what we have asked for is a "White Paper."  The very term, Conference Paper, generated a mini-storm about requirements, committees, academic standards, etc, etc, etc.  What we are asking to be submitted is a "White Paper" that may or may not contain references.  It's nice if they do but it's OK if they don't.  The Approval Committee is very sparse but, for now, consists of several "leaders" in this field who are not employed by a major corporation such as ILOG or Fair Isaac.

The LAST thing that we need is to be an academic conference OR to be a business meeting.  The former is far too stilted and dry while the latter is usually void of technical substance.  We have invited the very best of the very best of the very best (the creme de la creme de la creme) to participate and most of them have accepted.  Binding up the conference papers into a single collection that would be free for attendees and speakers and available for a minimal sum to others is not a big deal and we do NOT want to subject either the speakers nor the approval committee to "normal" academic conference mishmash and rigamarole.  On the other hand, we DO want the papers to be technical and address some of the issues outlined in the former blog on the call for White Papers.

Like some of you, I too have submitted (and had published) academic conference papers.  It's a pain to do the research, document almost every line with references, and your sole contribution is only about 10% of the total document.  For THIS conference, I would rather that the references and history part of the paper just be the list of references in the back of the paper and take up only 10% of the up-front text of the whole paper.  What we want to know is what are YOU presenting?  What is this new and innovative idea that YOU have and that you want to share?  What technically challenging idea did YOU have and  for which YOU devised a solution and YOU are presenting that solution?  In the words of an old Burger King commercial, "Where's the BEEF ?!!"  The conference is all about YOU and what YOU HAVE DONE - not about history and references and committees.

Maybe we will start something new in that we will have presentations that are exciting, technical, innovative and revolutionary and we will be sharing this with the wide spectrum of the  community of technical users.  The one continuing question from ORF 2008 (from the attendees) was that they want more presentations on HOW TO SOLVE THORNY PROBLEMS using a rulebase technology.  If you are presenting a new innovation, please tell us enough about it so that we can understand it - not just show results but explain HOW these results were made possible.  Also, remember, this conference in NOT just about rulebased systems and BRMS.  It's also about Prolog, LISP, AI, Constraint Based Programming, Complex Event Processing, Validation and Verification, Forecasting, Architecture, Modeling and lots more!!

Remember, Tutorials are on Monday from 8:00 a.m. to 6:00 p.m.  Newbies and Technical Business Managers (CIO, CTO, Project Managers, etc) will find that extremely helpful in understanding the rest of the conference.  This means that the conference presentation does NOT have to subscribe to the theory of presenting to the "lowest common denominator" of the attendees but to the architects and implementors of rulebased systems.

Most of all, we are there to have fun!  Vendors have new things to show off.   Architects and implementors have discovered new ways to solve complicated problems.  Scientists and researchers have new innovations to talk about.  And the attendees will be there to soak up every little thing and learn more in 3.5 days than they might learn all year from pouring through manuals and watching web casts.

SDG
jco

Monday, February 9, 2009

New Algorithm From Dr. Forgy

Greetings:

A brief discussion of the new algorithm is posted at http://javarules.blogspot.com/2009/02/new-algorithm-from-dr-forgy.html for your review and/or comments.  Of course, comments here should be viewable by more persons than that on my personal blog spot - especially since this blog has about 25 or 30 editors who can blog directly rather than just make comments.

Enjoy...

SDG
jco

Wednesday, February 4, 2009

ORF 2009 - Call for White Papers

Greetings:

Just a reminder that we need ALL proposed white paper abstracts in NOT LATER THAN 15 April 2009. However, the earlier that the abstracts are submitted, the easier will be our job since our "Approval Committee" is scattered around the world. Abstracts will be reviewed for

Content
Applicability
Technical focus
Potential interest to attendees

Possible subjects are:
Innovations to Rulebased Technology
Solutions to Technical Problems in Rulebased Technology
Homeland Security (problems with many object / many rules)
Processing Plants (Decisions on cracking tower problems)
Oil Rig or Building Design using Constraints
Shift Schedules using Constraint Based Programming
Scheduling Aircrafts, Trains, Taxis, Buses, etc.
Product Catalog Management
Customer Relationship Management (CRM)
Logistics (FedEx, UPS, etc.)
Configuration Problems (Aircraft, Computers, etc)
Complex Event Processing (Sliding Window Problem)
Rulebased Forecasting [my talk... :-) ]
Who gets a bailout and who doesn't :-)

Remember, this is a "sharing" conference where technical guys tell other technical guys HOW they did something.  We have seen plenty of PPTs about WHAT someone did, but we want to know how they did it.   For example, if you had to architect 20K rules and several thousand objects, how did you determine HOW to break out the rules into the various rule sets.  If you wrote something that utilized CEP then tell us HOW you solved the CEP problem.  What I'm going to share is the rational behind the RBF process, including the statistics, theory, etc.  Almost any business person should be really interested in reviewing their forecasting theory and statical classes.  :-)

Approvals for the abstracts will be sent out at the end of April or before so that you can get started on the white paper itself. At the end of the conference, the white papers will be collected into a Conference Proceedings. Whether they will be sold as bound copies (very expensive) or as electronic versions (no security in this one) we have not decided at this point. Probably, we will wait until we get the abstracts in before we make any kind of decision on that front.

The white paper itself must be submitted NOT LATER THAN 15 JULY 2009. Again, the earlier the better. The overhead that supports that goes along with the white paper must be submitted NOT LATER THAN 30 AUGUST 2009 in order to have it ready for the attendees.

At this point permit me to point out that we are not trying to be Geeks talking to Geeks about Geek things and nobody else should attend. Not at all! If you are a CIO, CTO, CEO or Senior Business Analysts then OF COURSE you are most welcome at October Rules Fest!  Especially if you want to learn about the "internals" of how a rulebased system actually works - something that I wish all business guys would do.  (Also, it's a great place to meet the Uber Geeks of the Rulebased Industry if you are looking for someone to help with your projects.)

SDG
jco