I'm sure you know how LEX and MEA work, so the following explication is for the benefit of the wider mailing list audience:
In the LEX conflict resolution strategy, rule activations are ordered by recency (as in the current Jess "depth" strategy). Among activations with the same recency, activations are ordered according to the specificity of their condition elements. The most recent rule activation with the most specific left-hand side (LHS) is fired first. Specificity refers to the total number of tests within all of the conditions of a rule. The MEA strategy (Means-Ends Analysis) has the same recency and specificity conditions as LEX, but places extra emphasis on the recency of a working memory element that matches the first condition element of a rule's LHS. MEA is used to facilitate the orderly handling of subgoals. If the first condition element of a rule is always a goal element, then the system will never be distracted by a very recent working memory element that is not a goal. (Brownston, Farrell, Kant, and Martin, Programming Expert Systems in OPS5, 1985(!?)) So presumably, the LEX strategy is intended to be used in systems where the orderly handling of subgoals is not needed or would get in the way of a more "organic" rule organization -- all condition elements in a set of rules having the same "weight" relative to each other. I could think of a number of domains where one could prefer the looser LEX strategy over MEA, for example an (object) classification system and a plan recognition system. In both cases, one possible approach is to let the entire LHS of some set of rules represent the pattern being recognized, without adornment of a goal element. Of course, the same domains could be implemented using MEA too. It all boils down to programming style and preferences. -Steve Solomon -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of James C. Owen Sent: Monday, March 13, 2006 2:27 PM To: [email protected] Subject: Re: JESS: Announcing Jess 7.0b6 OK, now I just "have" to know; why would anyone prefer LEX to MEA (Means Ends Analysis) for conflict resolution? I thought that Dr. Newell et al had reported on this quite some time ago and they showed why MEA was preferable to LEX. Like, the early 80's? So, again, why? (Not being argumentative - just looking for information...) SDG jco On Mar 13, 2006, at 12:20 PM, Steve Solomon wrote: > Is there any update on when the LEX conflict resolution strategy might > be implemented? > > Thanks, > > Steve Solomon > USC Institute for Creative Technologies > [EMAIL PROTECTED] > > > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] > On Behalf Of [EMAIL PROTECTED] > Sent: Sunday, March 12, 2006 6:39 PM > To: Jess Mailing List > Subject: JESS: Announcing Jess 7.0b6 > > Hi Folks, > > Version 7.0b6 of Jess, the Java Programmer's Rule Engine, is available > for download at the usual place: > > http://www.jessrules.com/download.shtml > > This version includes substantial documentation updates and some very > nice usability enhancements, including "Java Patterns" -- a new syntax > for writing simple slot tests that is more readable and natural for > Java programmers to use. From the change log: > > Fix bug in (logical) when modifying working memory from > nested scopes (thanks Yuri Gribov). Added "add" > function. Added "Java patterns." Deprecated > executeCommand() in favor of eval(). Massive > manual rewrites. Another logical/slot-specific bug > fixed (thanks Shan Ming Woo.) Rete member in > ClassSource transient (thanks Jonathan Sewall.) > Overloaded updateObject() accepting property name (for > Lakshmi Vempati). Fix synchronization of "undefinstance > *" (thanks Abdul Quddoos Khan). Slight change to "accumulate" > semantics (thanks Shan Ming Woo.) Overloaded "add()" > methods in ValueVector. > > The final release of Jess 7 is getting very close. As usual, comments > and questions are welcome on the mailing list at [email protected] > or directly to me at [EMAIL PROTECTED] Your feedback now is more > important than ever. > > --------------------------------------------------------- > Ernest Friedman-Hill > Advanced Software Research Phone: (925) 294-2154 > Sandia National Labs FAX: (925) 294-2234 > PO Box 969, MS 9012 [EMAIL PROTECTED] > Livermore, CA 94550 http://herzberg.ca.sandia.gov > > -------------------------------------------------------------------- > To unsubscribe, send the words 'unsubscribe jess-users > [EMAIL PROTECTED]' > in the BODY of a message to [EMAIL PROTECTED], NOT to the list > (use your own address!) List problems? Notify > [EMAIL PROTECTED] > -------------------------------------------------------------------- > > > -------------------------------------------------------------------- > To unsubscribe, send the words 'unsubscribe jess-users > [EMAIL PROTECTED]' > in the BODY of a message to [EMAIL PROTECTED], NOT to the list > (use your own address!) List problems? Notify owner-jess- > [EMAIL PROTECTED] > -------------------------------------------------------------------- > -------------------------------------------------------------------- To unsubscribe, send the words 'unsubscribe jess-users [EMAIL PROTECTED]' in the BODY of a message to [EMAIL PROTECTED], NOT to the list (use your own address!) List problems? Notify [EMAIL PROTECTED] -------------------------------------------------------------------- -------------------------------------------------------------------- To unsubscribe, send the words 'unsubscribe jess-users [EMAIL PROTECTED]' in the BODY of a message to [EMAIL PROTECTED], NOT to the list (use your own address!) List problems? Notify [EMAIL PROTECTED] --------------------------------------------------------------------
