aminePlatform.engines.memoryBasedInferences
Class MemoryBasedInferences
java.lang.Object
aminePlatform.engines.memoryBasedInferences.MemoryBasedInferences
- Direct Known Subclasses:
- MemoryAbductiveInference, MemoryAnalogicalInference, MemoryDeductiveInference
- public abstract class MemoryBasedInferences
- extends java.lang.Object
This class provides the possibility for deductive, abductive and analogical inferences,
starting from a given situation.
Copyright: Copyright (c) 2004-2009 Adil KABBAJ
Method Summary |
protected abstract CG |
doInferenceStep(CG newDescr,
SimulationObject simulationObject)
|
javax.swing.JTextArea |
getTextArea()
|
protected CG |
inferenceStep(CG newDescr)
inferenceStep represents one step in deductive/abductive inference: it asks for the position
of the given newDescr in the ontology/KB/memory, and check if newDescr is more specific than the antecedent
(for the case of deductive inference) or than the consequence (for the case of abductive inference) of an
existing CSRule, this later is activated; its antecedent/consequence is joined to newDescr and the result
of the maximalJoin is returned as an outcome of inferFrom. |
CG |
oneInferenceStep(CG newDescr)
oneStepInference represents one step in deductive/abductive inference: it asks for the position
of the given newDescr in the ontology/KB/memory, and check if newDescr is more specific than the antecedent
(for the case of deductive inference) or than the consequence (for the case of abductive inference) of an
existing CSRule, this later is activated; its antecedent/consequence is joined to newDescr and the result
of the maximalJoin is returned as an outcome of inferFrom. |
void |
setTextArea(javax.swing.JTextArea txtArea)
|
CG |
strictInferenceChain(CG newDescr)
strictInferenceChain corresponds to an iterative call to the oneStepInference method:
if it is found that the given newDescr is more specific than the antecedent (for the case
of deductive inference) or than the consequence (for the case of abductive inference) of an existing CSRule,
this later is activated; its antecedent/consequence is joined to newDescr. |
Methods inherited from class java.lang.Object |
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
DEDUCTIVE_INFERENCE_MODE
public static final byte DEDUCTIVE_INFERENCE_MODE
- See Also:
- Constant Field Values
ABDUCTIVE_INFERENCE_MODE
public static final byte ABDUCTIVE_INFERENCE_MODE
- See Also:
- Constant Field Values
ANALOGICAL_INFERENCE_MODE
public static final byte ANALOGICAL_INFERENCE_MODE
- See Also:
- Constant Field Values
ontology
Ontology ontology
lexicon
Lexicon lexicon
rulesUsed
AmineList rulesUsed
integrationSituation
public IntegrateSituation integrationSituation
txtArea
protected javax.swing.JTextArea txtArea
MemoryBasedInferences
public MemoryBasedInferences(Ontology ontology,
Lexicon lexicon)
throws DynamicKnowledgeIntegrationException
getTextArea
public javax.swing.JTextArea getTextArea()
setTextArea
public void setTextArea(javax.swing.JTextArea txtArea)
oneInferenceStep
public CG oneInferenceStep(CG newDescr)
throws java.lang.Exception
- oneStepInference represents one step in deductive/abductive inference: it asks for the position
of the given newDescr in the ontology/KB/memory, and check if newDescr is more specific than the antecedent
(for the case of deductive inference) or than the consequence (for the case of abductive inference) of an
existing CSRule, this later is activated; its antecedent/consequence is joined to newDescr and the result
of the maximalJoin is returned as an outcome of inferFrom. Otherwise, null is returned.
- Parameters:
newDescr
- CG
- Returns:
- CG
- Throws:
java.lang.Exception
strictInferenceChain
public CG strictInferenceChain(CG newDescr)
throws java.lang.Exception
- strictInferenceChain corresponds to an iterative call to the oneStepInference method:
if it is found that the given newDescr is more specific than the antecedent (for the case
of deductive inference) or than the consequence (for the case of abductive inference) of an existing CSRule,
this later is activated; its antecedent/consequence is joined to newDescr. The result of this
maximalJoin will be considered as the new value of newDescr and it will be integrated in turn, etc.
Note that we memorize the CSRules that have been activated so that each CSRule is activated only once,
to avoid cyclic activation of the same CSRule.
This method is called strict because the same kind of inference step (deductive or abductive) is performed along
the chain of inference; it is not allowed to perform deductive and abductive inference steps in the same chain.
- Parameters:
newDescr
- CG
- Returns:
- CG
- Throws:
java.lang.Exception
inferenceStep
protected CG inferenceStep(CG newDescr)
throws java.lang.Exception
- inferenceStep represents one step in deductive/abductive inference: it asks for the position
of the given newDescr in the ontology/KB/memory, and check if newDescr is more specific than the antecedent
(for the case of deductive inference) or than the consequence (for the case of abductive inference) of an
existing CSRule, this later is activated; its antecedent/consequence is joined to newDescr and the result
of the maximalJoin is returned as an outcome of inferFrom. Otherwise, null is returned.
- Parameters:
newDescr
- CG
- Returns:
- CG
- Throws:
java.lang.Exception
doInferenceStep
protected abstract CG doInferenceStep(CG newDescr,
SimulationObject simulationObject)
throws java.lang.Exception
- Throws:
java.lang.Exception