Re: JESS: question on slot specific
Argh, sorry, I used "no-loop" in my example, but I meant "slot- specific." On Jun 12, 2011, at 11:12 PM, Ernest Friedman-Hill wrote: Oooh, that's a bad example; I'll have to fix the manual. If you type (watch all) before your code, you'll see that the rule actually fires twice; once when B is nil, and once when it's 3. But the second time it fires, the "modify" has no effect, since the slot is being modified to contain the value it already contains. Try this rule instead, both with and without no-loop (and use (watch all) so you can see what's going on:) (defrule R (declare (no-loop TRUE)) ?d <- (D (A 1)) => (modify ?d (B (+ ?d.B 1 Note that your initial assert needs to have a number in the B slot, or you'll get an error. On Jun 12, 2011, at 8:08 PM, Felix Chan wrote: Hi, I have been trying to understand the slot-specific functionality which is described in the JESS manual as follows: --- 6.18. The 'slot-specific' declaration for deftemplates Deftemplate definitions can now include a "declare" section just as defrules can. There are several different properties that can be declared. One is "slot-specific". A template with this declaration will be matched in a special way: if a fact, created from such a template, which matches the left-hand-side of a rule is modified, the result depends on whether the modified slot is named in the pattern used to match the fact. As an example, consider the following: Jess> (deftemplate D (declare (slot-specific TRUE)) (slot A) (slot B)) Jess> (defrule R ?d <- (D (A 1)) => (modify ?d (B 3))) Without the "slot-specific" declaration, this rule would enter an endless loop, because it modifies a fact matched on the LHS in such a way that the modified fact will still match. With the declaration, it can simply fire once. This behavior is actually what many new users expect as the default; the technical term for it is refraction. -- I tried the following without the slot-specific declaration, expecting an infinite loop to occur: (reset) (deftemplate D (slot A) (slot B)) (defrule R ?d <- (D (A 1)) => (modify ?d (B 3))) (bind ?f (assert (D (A 1 (facts) (run) (facts) But I got the following outcome with no infinite loop: f-0 (MAIN::initial-fact) f-1 (MAIN::D (A 1) (B nil)) For a total of 2 facts in module MAIN. f-0 (MAIN::initial-fact) f-1 (MAIN::D (A 1) (B 3)) For a total of 2 facts in module MAIN. It seems slot-specific makes no difference in the outcome. I suspect I am missing a key concept here. Could someone explain? Thanks. Felix To unsubscribe, send the words 'unsubscribe jess-users y...@address.com ' in the BODY of a message to majord...@sandia.gov, NOT to the list (use your own address!) List problems? Notify owner-jess-us...@sandia.gov . - Ernest Friedman-Hill Informatics & Decision Sciences, Sandia National Laboratories PO Box 969, MS 9012, Livermore, CA 94550 http://www.jessrules.com To unsubscribe, send the words 'unsubscribe jess-users y...@address.com' in the BODY of a message to majord...@sandia.gov, NOT to the list (use your own address!) List problems? Notify owner-jess-us...@sandia.gov . - Ernest Friedman-Hill Informatics & Decision Sciences, Sandia National Laboratories PO Box 969, MS 9012, Livermore, CA 94550 http://www.jessrules.com To unsubscribe, send the words 'unsubscribe jess-users y...@address.com' in the BODY of a message to majord...@sandia.gov, NOT to the list (use your own address!) List problems? Notify owner-jess-us...@sandia.gov.
Re: JESS: question on slot specific
Oooh, that's a bad example; I'll have to fix the manual. If you type (watch all) before your code, you'll see that the rule actually fires twice; once when B is nil, and once when it's 3. But the second time it fires, the "modify" has no effect, since the slot is being modified to contain the value it already contains. Try this rule instead, both with and without no-loop (and use (watch all) so you can see what's going on:) (defrule R (declare (no-loop TRUE)) ?d <- (D (A 1)) => (modify ?d (B (+ ?d.B 1 Note that your initial assert needs to have a number in the B slot, or you'll get an error. On Jun 12, 2011, at 8:08 PM, Felix Chan wrote: Hi, I have been trying to understand the slot-specific functionality which is described in the JESS manual as follows: --- 6.18. The 'slot-specific' declaration for deftemplates Deftemplate definitions can now include a "declare" section just as defrules can. There are several different properties that can be declared. One is "slot-specific". A template with this declaration will be matched in a special way: if a fact, created from such a template, which matches the left-hand-side of a rule is modified, the result depends on whether the modified slot is named in the pattern used to match the fact. As an example, consider the following: Jess> (deftemplate D (declare (slot-specific TRUE)) (slot A) (slot B)) Jess> (defrule R ?d <- (D (A 1)) => (modify ?d (B 3))) Without the "slot-specific" declaration, this rule would enter an endless loop, because it modifies a fact matched on the LHS in such a way that the modified fact will still match. With the declaration, it can simply fire once. This behavior is actually what many new users expect as the default; the technical term for it is refraction. -- I tried the following without the slot-specific declaration, expecting an infinite loop to occur: (reset) (deftemplate D (slot A) (slot B)) (defrule R ?d <- (D (A 1)) => (modify ?d (B 3))) (bind ?f (assert (D (A 1 (facts) (run) (facts) But I got the following outcome with no infinite loop: f-0 (MAIN::initial-fact) f-1 (MAIN::D (A 1) (B nil)) For a total of 2 facts in module MAIN. f-0 (MAIN::initial-fact) f-1 (MAIN::D (A 1) (B 3)) For a total of 2 facts in module MAIN. It seems slot-specific makes no difference in the outcome. I suspect I am missing a key concept here. Could someone explain? Thanks. Felix To unsubscribe, send the words 'unsubscribe jess-users y...@address.com' in the BODY of a message to majord...@sandia.gov, NOT to the list (use your own address!) List problems? Notify owner-jess-us...@sandia.gov . - Ernest Friedman-Hill Informatics & Decision Sciences, Sandia National Laboratories PO Box 969, MS 9012, Livermore, CA 94550 http://www.jessrules.com To unsubscribe, send the words 'unsubscribe jess-users y...@address.com' in the BODY of a message to majord...@sandia.gov, NOT to the list (use your own address!) List problems? Notify owner-jess-us...@sandia.gov.
JESS: question on slot specific
Hi, I have been trying to understand the slot-specific functionality which is described in the JESS manual as follows: --- 6.18. The 'slot-specific' declaration for deftemplates Deftemplate definitions can now include a "declare" section just as defrules can. There are several different properties that can be declared. One is "slot-specific". A template with this declaration will be matched in a special way: if a fact, created from such a template, which matches the left-hand-side of a rule is modified, the result depends on whether the modified slot is named in the pattern used to match the fact. As an example, consider the following: Jess> (deftemplate D (declare (slot-specific TRUE)) (slot A) (slot B)) Jess> (defrule R ?d <- (D (A 1)) => (modify ?d (B 3))) Without the "slot-specific" declaration, this rule would enter an endless loop, because it modifies a fact matched on the LHS in such a way that the modified fact will still match. With the declaration, it can simply fire once. This behavior is actually what many new users expect as the default; the technical term for it is refraction. -- I tried the following without the slot-specific declaration, expecting an infinite loop to occur: (reset) (deftemplate D (slot A) (slot B)) (defrule R ?d <- (D (A 1)) => (modify ?d (B 3))) (bind ?f (assert (D (A 1 (facts) (run) (facts) But I got the following outcome with no infinite loop: f-0 (MAIN::initial-fact) f-1 (MAIN::D (A 1) (B nil)) For a total of 2 facts in module MAIN. f-0 (MAIN::initial-fact) f-1 (MAIN::D (A 1) (B 3)) For a total of 2 facts in module MAIN. It seems slot-specific makes no difference in the outcome. I suspect I am missing a key concept here. Could someone explain? Thanks. Felix To unsubscribe, send the words 'unsubscribe jess-users y...@address.com' in the BODY of a message to majord...@sandia.gov, NOT to the list (use your own address!) List problems? Notify owner-jess-us...@sandia.gov.
RE: JESS: On the Performance of Logical Retractions
If truth maintenance is a central part of your architecture, I recommend Building Problem Solvers, by Kenneth Forbus and Johan de Kleer. It's on Amazon: http://www.amazon.com/Building-Problem-Solvers-Artificial-Intelligence/dp/02 62061570/ref=sr_1_1?ie=UTF8&qid=1307815663&sr=8-1 and you can find the source code for the truth maintenance systems described in the book here: http://www.qrg.northwestern.edu/BPS/readme.html As part of my PhD work, I developed a reasoning system based on the LTRE, a forward-chaining rule engine on top of a logic-based TMS that is described in Building Problem Solvers. Coming from this background, I continually find Jess to be a Swiss Army knife of capabilities. However, if the logical conditional in Jess is not sufficient for your architecture, you'll probably need to implement a separate TMS layer. The logic-based TMS, which does fast (but incomplete) Boolean constraint propagation, provides a good balance between expressivity and efficiency. The problem solver architectures presented in Building Problem Solvers use the rule engine's rules to construct a problem-specific dependency network, through which the TMS propagates truth values. For example, the CyclePad system http://www.qrg.northwestern.edu/projects/NSF/Cyclepad/aboutcp.html enables the user to assemble and analyze thermodynamic cycles from a palette of devices (turbines, pumps, heaters, throttles, coolers, etc). Once the user has completed the cycle design, CyclePad runs its knowledge base of rules to generate a dependency network that captures the relationships among the thermodynamic properties at the inlet and outlet of each device. The user can choose the working fluid for the system, and this imposes further logical dependencies. For example, water will condense at certain combinations of pressure and temperature. The user analyzes the system by making assumptions about thermodynamic properties that the system then propagates through the dependency network. -John -Original Message- From: Ernest Friedman-Hill [mailto:ejfr...@sandia.gov] Sent: Saturday, June 11, 2011 8:20 AM To: jess-users@sandia.gov Subject: Re: JESS: On the Performance of Logical Retractions On Jun 11, 2011, at 6:11 AM, Oliya wrote: > > But still I have a question: what type of truth maintenance is > supported in Jess? Can you provide links to more information please. The "logical" conditional element is the only form of truth maintenance in Jess. I thought you said you were already using it? > Ernest Friedman-Hill Informatics & Decision Sciences, Sandia National Laboratories PO Box 969, MS 9012, Livermore, CA 94550 http://www.jessrules.com To unsubscribe, send the words 'unsubscribe jess-users y...@address.com' in the BODY of a message to majord...@sandia.gov, NOT to the list (use your own address!) List problems? Notify owner-jess-us...@sandia.gov. To unsubscribe, send the words 'unsubscribe jess-users y...@address.com' in the BODY of a message to majord...@sandia.gov, NOT to the list (use your own address!) List problems? Notify owner-jess-us...@sandia.gov.
