Menu
▾
▴
[DL-Learner SVN] SF.net SVN: dl-learner:[1608] trunk
|
From: <ku...@us...> - 2009-02-17 19:18:36
|
Revision: 1608
http://dl-learner.svn.sourceforge.net/dl-learner/?rev=1608&view=rev
Author: kurzum
Date: 2009-02-17 19:18:32 +0000 (Tue, 17 Feb 2009)
Log Message:
-----------
tried to fix null in configoptions.txt, a little bit better now, but not perfect. import(""); is a CLI anomaly.
fixed test script for extraction component
worked on manual
Modified Paths:
--------------
trunk/doc/configOptions.txt
trunk/doc/manual/manual.tex
trunk/src/dl-learner/org/dllearner/core/ComponentManager.java
trunk/src/dl-learner/org/dllearner/test/SparqlExtractionTest.java
Modified: trunk/doc/configOptions.txt
===================================================================
--- trunk/doc/configOptions.txt 2009-02-17 15:28:41 UTC (rev 1607)
+++ trunk/doc/configOptions.txt 2009-02-17 19:18:32 UTC (rev 1608)
@@ -7,7 +7,7 @@
component: KB file (org.dllearner.kb.KBFile)
============================================
-CLI usage: null = null;
+CLI usage: import ("$url", "null");
option name: url
description: URL pointer to the KB file
@@ -19,13 +19,13 @@
component: OWL API Ontology (org.dllearner.kb.OWLAPIOntology)
=============================================================
-CLI usage: null = null;
+CLI usage: import ("$url", "null");
component: OWL file (org.dllearner.kb.OWLFile)
==============================================
-CLI usage: null = owlfile;
+CLI usage: import ("$url", "owlfile");
option name: url
description: URL pointing to the OWL file
@@ -37,7 +37,7 @@
component: SPARQL Endpoint (org.dllearner.kb.sparql.SparqlKnowledgeSource)
==========================================================================
-CLI usage: null = sparql;
+CLI usage: import ("$url", "sparql");
option name: url
description: URL of SPARQL Endpoint
@@ -185,7 +185,7 @@
component: DIG reasoner (org.dllearner.reasoning.DIGReasoner)
=============================================================
-CLI usage: null = digReasoner;
+CLI usage: reasoner = digReasoner;
option name: reasonerUrl
description: URL of the DIG reasoner
@@ -209,7 +209,7 @@
component: fast instance checker (org.dllearner.reasoning.FastInstanceChecker)
==============================================================================
-CLI usage: null = fastInstanceChecker;
+CLI usage: reasoner = fastInstanceChecker;
option name: reasonerType
description: FaCT++ or Pellet to dematerialize
@@ -223,17 +223,23 @@
default value: true
CLI usage: fastInstanceChecker.defaultNegation = true;
+option name: forallRetrievalSemantics
+description: This option controls how to interpret the all quantifier in orall r.C. The standard option isto return all those which do not have an r-filler not in C. The domain semantics is to use thosewhich are in the domain of r and do not have an r-filler not in C. The forallExists semantics is touse those which have at least one r-filler and do not have an r-filler not in C.
+values: class org.dllearner.core.options.StringConfigOption
+default value: forallExists
+CLI usage: fastInstanceChecker.forallRetrievalSemantics = forallExists;
+
component: fast retrieval reasoner (org.dllearner.reasoning.FastRetrievalReasoner)
==================================================================================
-CLI usage: null = fastRetrievalReasoner;
+CLI usage: reasoner = fastRetrievalReasoner;
component: OWL API reasoner (org.dllearner.reasoning.OWLAPIReasoner)
====================================================================
-CLI usage: null = owlAPIReasoner;
+CLI usage: reasoner = owlAPIReasoner;
option name: reasonerType
description: FaCT++ or Pellet, which means "pellet" or "fact"
@@ -249,164 +255,152 @@
component: unnamed component (org.dllearner.learningproblems.ClassLearningProblem)
==================================================================================
-CLI usage: null = null;
+CLI usage: problem = classLearning;
option name: classToDescribe
description: class of which a description should be learned
values: class org.dllearner.core.options.StringConfigOption
default value: null
-CLI usage: null.classToDescribe = null;
+CLI usage: classLearning.classToDescribe = null;
option name: type
description: Whether to learn an equivalence class or super class axiom.
values: class org.dllearner.core.options.StringConfigOption
default value: equivalence
-CLI usage: null.type = equivalence;
+CLI usage: classLearning.type = equivalence;
-component: two valued definition learning problem (org.dllearner.learningproblems.PosNegDefinitionLP)
-=====================================================================================================
+component: two valued inclusion learning problem (org.dllearner.learningproblems.PosNegInclusionLP)
+===================================================================================================
-CLI usage: null = posNegDefinitionLP;
+CLI usage: problem = posNegInclusionLP;
option name: positiveExamples
description: positive examples
values: class org.dllearner.core.options.StringSetConfigOption
default value: null
-CLI usage: posNegDefinitionLP.positiveExamples = null;
+CLI usage: posNegInclusionLP.positiveExamples = null;
option name: negativeExamples
description: negative examples
values: class org.dllearner.core.options.StringSetConfigOption
default value: null
-CLI usage: posNegDefinitionLP.negativeExamples = null;
+CLI usage: posNegInclusionLP.negativeExamples = null;
option name: useRetrievalForClassficiation
description: Specifies whether to use retrieval or instance checks for testing a concept.
values: class org.dllearner.core.options.BooleanConfigOption
default value: false
-CLI usage: posNegDefinitionLP.useRetrievalForClassficiation = false;
+CLI usage: posNegInclusionLP.useRetrievalForClassficiation = false;
option name: percentPerLenghtUnit
description: describes the reduction in classification accuracy in percent one is willing to accept for reducing the length of the concept by one
values: class org.dllearner.core.options.DoubleConfigOption min 0.0 max 1.0
default value: 0.05
-CLI usage: posNegDefinitionLP.percentPerLenghtUnit = 0.05;
+CLI usage: posNegInclusionLP.percentPerLenghtUnit = 0.05;
option name: useMultiInstanceChecks
description: See UseMultiInstanceChecks enum.
values: class org.dllearner.core.options.StringConfigOption
default value: twoChecks
-CLI usage: posNegDefinitionLP.useMultiInstanceChecks = twoChecks;
+CLI usage: posNegInclusionLP.useMultiInstanceChecks = twoChecks;
-component: three valued definition learning problem (org.dllearner.learningproblems.PosNegDefinitionLPStrict)
-=============================================================================================================
+component: two valued definition learning problem (org.dllearner.learningproblems.PosNegLPStandard)
+===================================================================================================
-CLI usage: null = null;
+CLI usage: problem = posNegDefinitionLP;
option name: positiveExamples
description: positive examples
values: class org.dllearner.core.options.StringSetConfigOption
default value: null
-CLI usage: null.positiveExamples = null;
+CLI usage: posNegDefinitionLP.positiveExamples = null;
option name: negativeExamples
description: negative examples
values: class org.dllearner.core.options.StringSetConfigOption
default value: null
-CLI usage: null.negativeExamples = null;
+CLI usage: posNegDefinitionLP.negativeExamples = null;
option name: useRetrievalForClassficiation
description: Specifies whether to use retrieval or instance checks for testing a concept.
values: class org.dllearner.core.options.BooleanConfigOption
default value: false
-CLI usage: null.useRetrievalForClassficiation = false;
+CLI usage: posNegDefinitionLP.useRetrievalForClassficiation = false;
option name: percentPerLenghtUnit
description: describes the reduction in classification accuracy in percent one is willing to accept for reducing the length of the concept by one
values: class org.dllearner.core.options.DoubleConfigOption min 0.0 max 1.0
default value: 0.05
-CLI usage: null.percentPerLenghtUnit = 0.05;
+CLI usage: posNegDefinitionLP.percentPerLenghtUnit = 0.05;
option name: useMultiInstanceChecks
description: See UseMultiInstanceChecks enum.
values: class org.dllearner.core.options.StringConfigOption
default value: twoChecks
-CLI usage: null.useMultiInstanceChecks = twoChecks;
+CLI usage: posNegDefinitionLP.useMultiInstanceChecks = twoChecks;
-option name: penaliseNeutralExamples
-description: if set to true neutral examples are penalised
-values: class org.dllearner.core.options.BooleanConfigOption
-default value: null
-CLI usage: null.penaliseNeutralExamples = null;
-option name: accuracyPenalty
-description: penalty for pos/neg examples which are classified as neutral
-values: class org.dllearner.core.options.DoubleConfigOption
-default value: 1.0
-CLI usage: null.accuracyPenalty = 1.0;
-
-option name: errorPenalty
-description: penalty for pos. examples classified as negative or vice versa
-values: class org.dllearner.core.options.DoubleConfigOption
-default value: 3.0
-CLI usage: null.errorPenalty = 3.0;
-
-
-component: two valued inclusion learning problem (org.dllearner.learningproblems.PosNegInclusionLP)
+component: three valued definition learning problem (org.dllearner.learningproblems.PosNegLPStrict)
===================================================================================================
-CLI usage: null = posNegInclusionLP;
+CLI usage: problem = posNegDefinitionLPStrict;
option name: positiveExamples
description: positive examples
values: class org.dllearner.core.options.StringSetConfigOption
default value: null
-CLI usage: posNegInclusionLP.positiveExamples = null;
+CLI usage: posNegDefinitionLPStrict.positiveExamples = null;
option name: negativeExamples
description: negative examples
values: class org.dllearner.core.options.StringSetConfigOption
default value: null
-CLI usage: posNegInclusionLP.negativeExamples = null;
+CLI usage: posNegDefinitionLPStrict.negativeExamples = null;
option name: useRetrievalForClassficiation
description: Specifies whether to use retrieval or instance checks for testing a concept.
values: class org.dllearner.core.options.BooleanConfigOption
default value: false
-CLI usage: posNegInclusionLP.useRetrievalForClassficiation = false;
+CLI usage: posNegDefinitionLPStrict.useRetrievalForClassficiation = false;
option name: percentPerLenghtUnit
description: describes the reduction in classification accuracy in percent one is willing to accept for reducing the length of the concept by one
values: class org.dllearner.core.options.DoubleConfigOption min 0.0 max 1.0
default value: 0.05
-CLI usage: posNegInclusionLP.percentPerLenghtUnit = 0.05;
+CLI usage: posNegDefinitionLPStrict.percentPerLenghtUnit = 0.05;
option name: useMultiInstanceChecks
description: See UseMultiInstanceChecks enum.
values: class org.dllearner.core.options.StringConfigOption
default value: twoChecks
-CLI usage: posNegInclusionLP.useMultiInstanceChecks = twoChecks;
+CLI usage: posNegDefinitionLPStrict.useMultiInstanceChecks = twoChecks;
+option name: penaliseNeutralExamples
+description: if set to true neutral examples are penalised
+values: class org.dllearner.core.options.BooleanConfigOption
+default value: null
+CLI usage: posNegDefinitionLPStrict.penaliseNeutralExamples = null;
-component: positive only definition learning problem (org.dllearner.learningproblems.PosOnlyDefinitionLP)
-=========================================================================================================
+option name: accuracyPenalty
+description: penalty for pos/neg examples which are classified as neutral
+values: class org.dllearner.core.options.DoubleConfigOption
+default value: 1.0
+CLI usage: posNegDefinitionLPStrict.accuracyPenalty = 1.0;
-CLI usage: null = posOnlyDefinitionLP;
+option name: errorPenalty
+description: penalty for pos. examples classified as negative or vice versa
+values: class org.dllearner.core.options.DoubleConfigOption
+default value: 3.0
+CLI usage: posNegDefinitionLPStrict.errorPenalty = 3.0;
-option name: positiveExamples
-description: positive examples
-values: class org.dllearner.core.options.StringSetConfigOption
-default value: null
-CLI usage: posOnlyDefinitionLP.positiveExamples = null;
+component: positive only definition learning problem (org.dllearner.learningproblems.PosOnlyLP)
+===============================================================================================
-component: pos only inclusion learning problem (org.dllearner.learningproblems.PosOnlyInclusionLP)
-==================================================================================================
+CLI usage: problem = null;
-CLI usage: null = null;
-
option name: positiveExamples
description: positive examples
values: class org.dllearner.core.options.StringSetConfigOption
@@ -421,7 +415,7 @@
component: brute force learning algorithm (org.dllearner.algorithms.BruteForceLearner)
======================================================================================
-CLI usage: null = bruteForce;
+CLI usage: algorithm = bruteForce;
option name: maxLength
description: maximum length of generated concepts
@@ -436,226 +430,112 @@
CLI usage: bruteForce.returnType = null;
-component: DBpedia Navigation Suggestor (org.dllearner.algorithms.DBpediaNavigationSuggestor)
-=============================================================================================
+component: random guesser learning algorithm (org.dllearner.algorithms.RandomGuesser)
+=====================================================================================
-CLI usage: null = dbpediaNavigationSuggestor;
+CLI usage: algorithm = random;
-option name: writeSearchTree
-description: specifies whether to write a search tree
-values: class org.dllearner.core.options.BooleanConfigOption
-default value: false
-CLI usage: dbpediaNavigationSuggestor.writeSearchTree = false;
+option name: numberOfTrees
+description: number of randomly generated concepts/trees
+values: class org.dllearner.core.options.IntegerConfigOption
+default value: 5
+CLI usage: random.numberOfTrees = 5;
-option name: searchTreeFile
-description: file to use for the search tree
-values: class org.dllearner.core.options.StringConfigOption
-default value: log/searchTree.txt
-CLI usage: dbpediaNavigationSuggestor.searchTreeFile = log/searchTree.txt;
+option name: maxDepth
+description: maximum depth of generated concepts/trees
+values: class org.dllearner.core.options.IntegerConfigOption
+default value: 5
+CLI usage: random.maxDepth = 5;
-option name: replaceSearchTree
-description: specifies whether to replace the search tree in the log file after each run or append the new search tree
-values: class org.dllearner.core.options.BooleanConfigOption
-default value: false
-CLI usage: dbpediaNavigationSuggestor.replaceSearchTree = false;
-option name: heuristic
-description: specifiy the heuristic to use
-values: class org.dllearner.core.options.StringConfigOption
-default value: lexicographic
-CLI usage: dbpediaNavigationSuggestor.heuristic = lexicographic;
+component: CELOE (org.dllearner.algorithms.celoe.CELOE)
+=======================================================
-option name: applyAllFilter
-description: usage of equivalence ALL R.C AND ALL R.D = ALL R.(C AND D)
-values: class org.dllearner.core.options.BooleanConfigOption
-default value: true
-CLI usage: dbpediaNavigationSuggestor.applyAllFilter = true;
+CLI usage: algorithm = celoe;
-option name: applyExistsFilter
-description: usage of equivalence EXISTS R.C OR EXISTS R.D = EXISTS R.(C OR D)
-values: class org.dllearner.core.options.BooleanConfigOption
-default value: true
-CLI usage: dbpediaNavigationSuggestor.applyExistsFilter = true;
-
-option name: useTooWeakList
-description: try to filter out too weak concepts without sending them to the reasoner
-values: class org.dllearner.core.options.BooleanConfigOption
-default value: true
-CLI usage: dbpediaNavigationSuggestor.useTooWeakList = true;
-
-option name: useOverlyGeneralList
-description: try to find overly general concept without sending them to the reasoner
-values: class org.dllearner.core.options.BooleanConfigOption
-default value: true
-CLI usage: dbpediaNavigationSuggestor.useOverlyGeneralList = true;
-
-option name: useShortConceptConstruction
-description: shorten concept to see whether they already exist
-values: class org.dllearner.core.options.BooleanConfigOption
-default value: true
-CLI usage: dbpediaNavigationSuggestor.useShortConceptConstruction = true;
-
-option name: horizontalExpansionFactor
-description: horizontal expansion factor (see publication for description)
-values: class org.dllearner.core.options.DoubleConfigOption min 0.0 max 1.0
-default value: 0.6
-CLI usage: dbpediaNavigationSuggestor.horizontalExpansionFactor = 0.6;
-
-option name: improveSubsumptionHierarchy
-description: simplify subsumption hierarchy to reduce search space (see publication for description)
-values: class org.dllearner.core.options.BooleanConfigOption
-default value: true
-CLI usage: dbpediaNavigationSuggestor.improveSubsumptionHierarchy = true;
-
-option name: allowedConcepts
-description: concepts the algorithm is allowed to use
-values: class org.dllearner.core.options.StringSetConfigOption
-default value: null
-CLI usage: dbpediaNavigationSuggestor.allowedConcepts = null;
-
-option name: ignoredConcepts
-description: concepts the algorithm must ignore
-values: class org.dllearner.core.options.StringSetConfigOption
-default value: null
-CLI usage: dbpediaNavigationSuggestor.ignoredConcepts = null;
-
-option name: allowedRoles
-description: roles the algorithm is allowed to use
-values: class org.dllearner.core.options.StringSetConfigOption
-default value: null
-CLI usage: dbpediaNavigationSuggestor.allowedRoles = null;
-
-option name: ignoredRoles
-description: roles the algorithm must ignore
-values: class org.dllearner.core.options.StringSetConfigOption
-default value: null
-CLI usage: dbpediaNavigationSuggestor.ignoredRoles = null;
-
option name: useAllConstructor
description: specifies whether the universal concept constructor is used in the learning algorithm
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: dbpediaNavigationSuggestor.useAllConstructor = true;
+CLI usage: celoe.useAllConstructor = true;
option name: useExistsConstructor
description: specifies whether the existential concept constructor is used in the learning algorithm
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: dbpediaNavigationSuggestor.useExistsConstructor = true;
+CLI usage: celoe.useExistsConstructor = true;
-option name: useCardinalityRestrictions
-description: specifies whether CardinalityRestrictions is used in the learning algorithm
-values: class org.dllearner.core.options.BooleanConfigOption
-default value: true
-CLI usage: dbpediaNavigationSuggestor.useCardinalityRestrictions = true;
-
option name: useHasValueConstructor
description: specifies whether the hasValue constructor is used in the learning algorithm
values: class org.dllearner.core.options.BooleanConfigOption
default value: false
-CLI usage: dbpediaNavigationSuggestor.useHasValueConstructor = false;
+CLI usage: celoe.useHasValueConstructor = false;
option name: valueFrequencyThreshold
description: specifies how often an object must occur as value in order to be considered for hasValue restrictions
values: class org.dllearner.core.options.IntegerConfigOption
default value: 3
-CLI usage: dbpediaNavigationSuggestor.valueFrequencyThreshold = 3;
+CLI usage: celoe.valueFrequencyThreshold = 3;
+option name: useCardinalityRestrictions
+description: specifies whether CardinalityRestrictions is used in the learning algorithm
+values: class org.dllearner.core.options.BooleanConfigOption
+default value: true
+CLI usage: celoe.useCardinalityRestrictions = true;
+
+option name: cardinalityLimit
+description: Gives the maximum number used in cardinality restrictions.
+values: class org.dllearner.core.options.IntegerConfigOption
+default value: 5
+CLI usage: celoe.cardinalityLimit = 5;
+
option name: useNegation
description: specifies whether negation is used in the learning algorothm
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: dbpediaNavigationSuggestor.useNegation = true;
+CLI usage: celoe.useNegation = true;
option name: useBooleanDatatypes
description: specifies whether boolean datatypes are used in the learning algorothm
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: dbpediaNavigationSuggestor.useBooleanDatatypes = true;
+CLI usage: celoe.useBooleanDatatypes = true;
+option name: useDoubleDatatypes
+description: specifies whether boolean datatypes are used in the learning algorothm
+values: class org.dllearner.core.options.BooleanConfigOption
+default value: true
+CLI usage: celoe.useDoubleDatatypes = true;
+
option name: maxExecutionTimeInSeconds
description: algorithm will stop after specified seconds
values: class org.dllearner.core.options.IntegerConfigOption
-default value: 0
-CLI usage: dbpediaNavigationSuggestor.maxExecutionTimeInSeconds = 0;
+default value: 10
+CLI usage: celoe.maxExecutionTimeInSeconds = 10;
-option name: minExecutionTimeInSeconds
-description: algorithm will run at least specified seconds
-values: class org.dllearner.core.options.IntegerConfigOption
-default value: 0
-CLI usage: dbpediaNavigationSuggestor.minExecutionTimeInSeconds = 0;
-
-option name: guaranteeXgoodDescriptions
-description: algorithm will run until X good (100%) concept descritpions are found
-values: class org.dllearner.core.options.IntegerConfigOption
-default value: 1
-CLI usage: dbpediaNavigationSuggestor.guaranteeXgoodDescriptions = 1;
-
-option name: logLevel
-description: determines the logLevel for this component, can be {TRACE, DEBUG, INFO}
-values: class org.dllearner.core.options.StringConfigOption
-default value: DEBUG
-CLI usage: dbpediaNavigationSuggestor.logLevel = DEBUG;
-
option name: noisePercentage
description: the (approximated) percentage of noise within the examples
values: class org.dllearner.core.options.DoubleConfigOption min 0.0 max 100.0
default value: 0.0
-CLI usage: dbpediaNavigationSuggestor.noisePercentage = 0.0;
+CLI usage: celoe.noisePercentage = 0.0;
-option name: startClass
-description: the named class which should be used to start the algorithm (GUI: needs a widget for selecting a class)
-values: class org.dllearner.core.options.StringConfigOption
-default value: null
-CLI usage: dbpediaNavigationSuggestor.startClass = null;
-
-option name: forceRefinementLengthIncrease
-description: specifies whether nodes should be expanded until only longer refinements are reached
-values: class org.dllearner.core.options.BooleanConfigOption
-default value: null
-CLI usage: dbpediaNavigationSuggestor.forceRefinementLengthIncrease = null;
-
-
-component: random guesser learning algorithm (org.dllearner.algorithms.RandomGuesser)
-=====================================================================================
-
-CLI usage: null = random;
-
-option name: numberOfTrees
-description: number of randomly generated concepts/trees
-values: class org.dllearner.core.options.IntegerConfigOption
-default value: 5
-CLI usage: random.numberOfTrees = 5;
-
option name: maxDepth
-description: maximum depth of generated concepts/trees
+description: maximum depth of description
values: class org.dllearner.core.options.IntegerConfigOption
-default value: 5
-CLI usage: random.maxDepth = 5;
+default value: 4
+CLI usage: celoe.maxDepth = 4;
-component: simple suggestion algorithm (org.dllearner.algorithms.SimpleSuggestionLearningAlgorithm)
-===================================================================================================
-
-CLI usage: null = null;
-
-
-component: unnamed component (org.dllearner.algorithms.celoe.CELOE)
-===================================================================
-
-CLI usage: null = null;
-
-
component: standard EL learning algorithm (org.dllearner.algorithms.el.ELLearningAlgorithm)
===========================================================================================
-CLI usage: null = el;
+CLI usage: algorithm = el;
component: genetic programming learning algorithm (org.dllearner.algorithms.gp.GP)
==================================================================================
-CLI usage: null = gp;
+CLI usage: algorithm = gp;
option name: selectionType
description: selection type
@@ -760,399 +640,399 @@
CLI usage: gp.maxConceptLength = 75;
-component: refinement operator based learning algorithm II (org.dllearner.algorithms.refexamples.ExampleBasedROLComponent)
-==========================================================================================================================
+component: refinement operator based learning algorithm (org.dllearner.algorithms.refinement.ROLearner)
+=======================================================================================================
-CLI usage: null = refexamples;
+CLI usage: algorithm = refinement;
option name: writeSearchTree
description: specifies whether to write a search tree
values: class org.dllearner.core.options.BooleanConfigOption
default value: false
-CLI usage: refexamples.writeSearchTree = false;
+CLI usage: refinement.writeSearchTree = false;
option name: searchTreeFile
description: file to use for the search tree
values: class org.dllearner.core.options.StringConfigOption
default value: log/searchTree.txt
-CLI usage: refexamples.searchTreeFile = log/searchTree.txt;
+CLI usage: refinement.searchTreeFile = log/searchTree.txt;
option name: replaceSearchTree
description: specifies whether to replace the search tree in the log file after each run or append the new search tree
values: class org.dllearner.core.options.BooleanConfigOption
default value: false
-CLI usage: refexamples.replaceSearchTree = false;
+CLI usage: refinement.replaceSearchTree = false;
option name: heuristic
description: specifiy the heuristic to use
values: class org.dllearner.core.options.StringConfigOption
default value: lexicographic
-CLI usage: refexamples.heuristic = lexicographic;
+CLI usage: refinement.heuristic = lexicographic;
option name: applyAllFilter
description: usage of equivalence ALL R.C AND ALL R.D = ALL R.(C AND D)
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refexamples.applyAllFilter = true;
+CLI usage: refinement.applyAllFilter = true;
option name: applyExistsFilter
description: usage of equivalence EXISTS R.C OR EXISTS R.D = EXISTS R.(C OR D)
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refexamples.applyExistsFilter = true;
+CLI usage: refinement.applyExistsFilter = true;
option name: useTooWeakList
description: try to filter out too weak concepts without sending them to the reasoner
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refexamples.useTooWeakList = true;
+CLI usage: refinement.useTooWeakList = true;
option name: useOverlyGeneralList
description: try to find overly general concept without sending them to the reasoner
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refexamples.useOverlyGeneralList = true;
+CLI usage: refinement.useOverlyGeneralList = true;
option name: useShortConceptConstruction
description: shorten concept to see whether they already exist
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refexamples.useShortConceptConstruction = true;
+CLI usage: refinement.useShortConceptConstruction = true;
option name: horizontalExpansionFactor
description: horizontal expansion factor (see publication for description)
values: class org.dllearner.core.options.DoubleConfigOption min 0.0 max 1.0
default value: 0.6
-CLI usage: refexamples.horizontalExpansionFactor = 0.6;
+CLI usage: refinement.horizontalExpansionFactor = 0.6;
option name: improveSubsumptionHierarchy
description: simplify subsumption hierarchy to reduce search space (see publication for description)
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refexamples.improveSubsumptionHierarchy = true;
+CLI usage: refinement.improveSubsumptionHierarchy = true;
+option name: quiet
+description: may be deprecated soon
+values: class org.dllearner.core.options.BooleanConfigOption
+default value: false
+CLI usage: refinement.quiet = false;
+
option name: allowedConcepts
description: concepts the algorithm is allowed to use
values: class org.dllearner.core.options.StringSetConfigOption
default value: null
-CLI usage: refexamples.allowedConcepts = null;
+CLI usage: refinement.allowedConcepts = null;
option name: ignoredConcepts
description: concepts the algorithm must ignore
values: class org.dllearner.core.options.StringSetConfigOption
default value: null
-CLI usage: refexamples.ignoredConcepts = null;
+CLI usage: refinement.ignoredConcepts = null;
option name: allowedRoles
description: roles the algorithm is allowed to use
values: class org.dllearner.core.options.StringSetConfigOption
default value: null
-CLI usage: refexamples.allowedRoles = null;
+CLI usage: refinement.allowedRoles = null;
option name: ignoredRoles
description: roles the algorithm must ignore
values: class org.dllearner.core.options.StringSetConfigOption
default value: null
-CLI usage: refexamples.ignoredRoles = null;
+CLI usage: refinement.ignoredRoles = null;
option name: useAllConstructor
description: specifies whether the universal concept constructor is used in the learning algorithm
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refexamples.useAllConstructor = true;
+CLI usage: refinement.useAllConstructor = true;
option name: useExistsConstructor
description: specifies whether the existential concept constructor is used in the learning algorithm
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refexamples.useExistsConstructor = true;
+CLI usage: refinement.useExistsConstructor = true;
-option name: useHasValueConstructor
-description: specifies whether the hasValue constructor is used in the learning algorithm
+option name: useNegation
+description: specifies whether negation is used in the learning algorothm
values: class org.dllearner.core.options.BooleanConfigOption
-default value: false
-CLI usage: refexamples.useHasValueConstructor = false;
+default value: true
+CLI usage: refinement.useNegation = true;
-option name: valueFrequencyThreshold
-description: specifies how often an object must occur as value in order to be considered for hasValue restrictions
-values: class org.dllearner.core.options.IntegerConfigOption
-default value: 3
-CLI usage: refexamples.valueFrequencyThreshold = 3;
-
option name: useCardinalityRestrictions
description: specifies whether CardinalityRestrictions is used in the learning algorithm
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refexamples.useCardinalityRestrictions = true;
+CLI usage: refinement.useCardinalityRestrictions = true;
-option name: cardinalityLimit
-description: Gives the maximum number used in cardinality restrictions.
-values: class org.dllearner.core.options.IntegerConfigOption
-default value: 5
-CLI usage: refexamples.cardinalityLimit = 5;
-
-option name: useNegation
-description: specifies whether negation is used in the learning algorothm
-values: class org.dllearner.core.options.BooleanConfigOption
-default value: true
-CLI usage: refexamples.useNegation = true;
-
option name: useBooleanDatatypes
description: specifies whether boolean datatypes are used in the learning algorothm
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refexamples.useBooleanDatatypes = true;
+CLI usage: refinement.useBooleanDatatypes = true;
-option name: useDoubleDatatypes
-description: specifies whether boolean datatypes are used in the learning algorothm
-values: class org.dllearner.core.options.BooleanConfigOption
-default value: true
-CLI usage: refexamples.useDoubleDatatypes = true;
-
option name: maxExecutionTimeInSeconds
description: algorithm will stop after specified seconds
values: class org.dllearner.core.options.IntegerConfigOption
default value: 0
-CLI usage: refexamples.maxExecutionTimeInSeconds = 0;
+CLI usage: refinement.maxExecutionTimeInSeconds = 0;
option name: minExecutionTimeInSeconds
description: algorithm will run at least specified seconds
values: class org.dllearner.core.options.IntegerConfigOption
default value: 0
-CLI usage: refexamples.minExecutionTimeInSeconds = 0;
+CLI usage: refinement.minExecutionTimeInSeconds = 0;
option name: guaranteeXgoodDescriptions
description: algorithm will run until X good (100%) concept descritpions are found
values: class org.dllearner.core.options.IntegerConfigOption
default value: 1
-CLI usage: refexamples.guaranteeXgoodDescriptions = 1;
+CLI usage: refinement.guaranteeXgoodDescriptions = 1;
-option name: maxClassDescriptionTests
-description: The maximum number of candidate hypothesis the algorithm is allowed to test (0 = no limit). The algorithm will stop afterwards. (The real number of tests can be slightly higher, because this criterion usually won't be checked after each single test.)
-values: class org.dllearner.core.options.IntegerConfigOption
-default value: 0
-CLI usage: refexamples.maxClassDescriptionTests = 0;
-
option name: logLevel
description: determines the logLevel for this component, can be {TRACE, DEBUG, INFO}
values: class org.dllearner.core.options.StringConfigOption
default value: DEBUG
-CLI usage: refexamples.logLevel = DEBUG;
+CLI usage: refinement.logLevel = DEBUG;
-option name: usePropernessChecks
-description: specifies whether to check for equivalence (i.e. discard equivalent refinements)
-values: class org.dllearner.core.options.BooleanConfigOption
-default value: false
-CLI usage: refexamples.usePropernessChecks = false;
-option name: maxPosOnlyExpansion
-description: specifies how often a node in the search tree of a posonly learning problem needs to be expanded before it is considered as solution candidate
-values: class org.dllearner.core.options.IntegerConfigOption
-default value: 4
-CLI usage: refexamples.maxPosOnlyExpansion = 4;
+component: refinement operator based learning algorithm II (org.dllearner.algorithms.refinement2.ROLComponent2)
+===============================================================================================================
-option name: noisePercentage
-description: the (approximated) percentage of noise within the examples
-values: class org.dllearner.core.options.DoubleConfigOption min 0.0 max 100.0
-default value: 0.0
-CLI usage: refexamples.noisePercentage = 0.0;
+CLI usage: algorithm = refexamples;
-option name: terminateOnNoiseReached
-description: specifies whether to terminate when noise criterion is met
-values: class org.dllearner.core.options.BooleanConfigOption
-default value: true
-CLI usage: refexamples.terminateOnNoiseReached = true;
-
-option name: startClass
-description: the named class which should be used to start the algorithm (GUI: needs a widget for selecting a class)
-values: class org.dllearner.core.options.StringConfigOption
-default value: null
-CLI usage: refexamples.startClass = null;
-
-option name: forceRefinementLengthIncrease
-description: specifies whether nodes should be expanded until only longer refinements are reached
-values: class org.dllearner.core.options.BooleanConfigOption
-default value: null
-CLI usage: refexamples.forceRefinementLengthIncrease = null;
-
-option name: negativeWeight
-description: Used to penalise errors on negative examples different from those of positive examples (lower = less importance for negatives).
-values: class org.dllearner.core.options.DoubleConfigOption
-default value: 1.0
-CLI usage: refexamples.negativeWeight = 1.0;
-
-option name: startNodeBonus
-description: You can use this to give a heuristic bonus on the start node (= initially broader exploration of search space).
-values: class org.dllearner.core.options.DoubleConfigOption
-default value: 0.0
-CLI usage: refexamples.startNodeBonus = 0.0;
-
-option name: negationPenalty
-description: Penalty on negations (TODO: better explanation).
-values: class org.dllearner.core.options.IntegerConfigOption
-default value: 0
-CLI usage: refexamples.negationPenalty = 0;
-
-option name: expansionPenaltyFactor
-description: describes the reduction in heuristic score one is willing to accept for reducing the length of the concept by one
-values: class org.dllearner.core.options.DoubleConfigOption
-default value: 0.02
-CLI usage: refexamples.expansionPenaltyFactor = 0.02;
-
-
-component: refinement operator based learning algorithm (org.dllearner.algorithms.refinement.ROLearner)
-=======================================================================================================
-
-CLI usage: null = refinement;
-
option name: writeSearchTree
description: specifies whether to write a search tree
values: class org.dllearner.core.options.BooleanConfigOption
default value: false
-CLI usage: refinement.writeSearchTree = false;
+CLI usage: refexamples.writeSearchTree = false;
option name: searchTreeFile
description: file to use for the search tree
values: class org.dllearner.core.options.StringConfigOption
default value: log/searchTree.txt
-CLI usage: refinement.searchTreeFile = log/searchTree.txt;
+CLI usage: refexamples.searchTreeFile = log/searchTree.txt;
option name: replaceSearchTree
description: specifies whether to replace the search tree in the log file after each run or append the new search tree
values: class org.dllearner.core.options.BooleanConfigOption
default value: false
-CLI usage: refinement.replaceSearchTree = false;
+CLI usage: refexamples.replaceSearchTree = false;
option name: heuristic
description: specifiy the heuristic to use
values: class org.dllearner.core.options.StringConfigOption
default value: lexicographic
-CLI usage: refinement.heuristic = lexicographic;
+CLI usage: refexamples.heuristic = lexicographic;
option name: applyAllFilter
description: usage of equivalence ALL R.C AND ALL R.D = ALL R.(C AND D)
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refinement.applyAllFilter = true;
+CLI usage: refexamples.applyAllFilter = true;
option name: applyExistsFilter
description: usage of equivalence EXISTS R.C OR EXISTS R.D = EXISTS R.(C OR D)
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refinement.applyExistsFilter = true;
+CLI usage: refexamples.applyExistsFilter = true;
option name: useTooWeakList
description: try to filter out too weak concepts without sending them to the reasoner
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refinement.useTooWeakList = true;
+CLI usage: refexamples.useTooWeakList = true;
option name: useOverlyGeneralList
description: try to find overly general concept without sending them to the reasoner
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refinement.useOverlyGeneralList = true;
+CLI usage: refexamples.useOverlyGeneralList = true;
option name: useShortConceptConstruction
description: shorten concept to see whether they already exist
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refinement.useShortConceptConstruction = true;
+CLI usage: refexamples.useShortConceptConstruction = true;
option name: horizontalExpansionFactor
description: horizontal expansion factor (see publication for description)
values: class org.dllearner.core.options.DoubleConfigOption min 0.0 max 1.0
default value: 0.6
-CLI usage: refinement.horizontalExpansionFactor = 0.6;
+CLI usage: refexamples.horizontalExpansionFactor = 0.6;
option name: improveSubsumptionHierarchy
description: simplify subsumption hierarchy to reduce search space (see publication for description)
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refinement.improveSubsumptionHierarchy = true;
+CLI usage: refexamples.improveSubsumptionHierarchy = true;
-option name: quiet
-description: may be deprecated soon
-values: class org.dllearner.core.options.BooleanConfigOption
-default value: false
-CLI usage: refinement.quiet = false;
-
option name: allowedConcepts
description: concepts the algorithm is allowed to use
values: class org.dllearner.core.options.StringSetConfigOption
default value: null
-CLI usage: refinement.allowedConcepts = null;
+CLI usage: refexamples.allowedConcepts = null;
option name: ignoredConcepts
description: concepts the algorithm must ignore
values: class org.dllearner.core.options.StringSetConfigOption
default value: null
-CLI usage: refinement.ignoredConcepts = null;
+CLI usage: refexamples.ignoredConcepts = null;
option name: allowedRoles
description: roles the algorithm is allowed to use
values: class org.dllearner.core.options.StringSetConfigOption
default value: null
-CLI usage: refinement.allowedRoles = null;
+CLI usage: refexamples.allowedRoles = null;
option name: ignoredRoles
description: roles the algorithm must ignore
values: class org.dllearner.core.options.StringSetConfigOption
default value: null
-CLI usage: refinement.ignoredRoles = null;
+CLI usage: refexamples.ignoredRoles = null;
option name: useAllConstructor
description: specifies whether the universal concept constructor is used in the learning algorithm
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refinement.useAllConstructor = true;
+CLI usage: refexamples.useAllConstructor = true;
option name: useExistsConstructor
description: specifies whether the existential concept constructor is used in the learning algorithm
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refinement.useExistsConstructor = true;
+CLI usage: refexamples.useExistsConstructor = true;
-option name: useNegation
-description: specifies whether negation is used in the learning algorothm
+option name: useHasValueConstructor
+description: specifies whether the hasValue constructor is used in the learning algorithm
values: class org.dllearner.core.options.BooleanConfigOption
-default value: true
-CLI usage: refinement.useNegation = true;
+default value: false
+CLI usage: refexamples.useHasValueConstructor = false;
+option name: valueFrequencyThreshold
+description: specifies how often an object must occur as value in order to be considered for hasValue restrictions
+values: class org.dllearner.core.options.IntegerConfigOption
+default value: 3
+CLI usage: refexamples.valueFrequencyThreshold = 3;
+
option name: useCardinalityRestrictions
description: specifies whether CardinalityRestrictions is used in the learning algorithm
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refinement.useCardinalityRestrictions = true;
+CLI usage: refexamples.useCardinalityRestrictions = true;
+option name: cardinalityLimit
+description: Gives the maximum number used in cardinality restrictions.
+values: class org.dllearner.core.options.IntegerConfigOption
+default value: 5
+CLI usage: refexamples.cardinalityLimit = 5;
+
+option name: useNegation
+description: specifies whether negation is used in the learning algorothm
+values: class org.dllearner.core.options.BooleanConfigOption
+default value: true
+CLI usage: refexamples.useNegation = true;
+
option name: useBooleanDatatypes
description: specifies whether boolean datatypes are used in the learning algorothm
values: class org.dllearner.core.options.BooleanConfigOption
default value: true
-CLI usage: refinement.useBooleanDatatypes = true;
+CLI usage: refexamples.useBooleanDatatypes = true;
+option name: useDoubleDatatypes
+description: specifies whether boolean datatypes are used in the learning algorothm
+values: class org.dllearner.core.options.BooleanConfigOption
+default value: true
+CLI usage: refexamples.useDoubleDatatypes = true;
+
option name: maxExecutionTimeInSeconds
description: algorithm will stop after specified seconds
values: class org.dllearner.core.options.IntegerConfigOption
default value: 0
-CLI usage: refinement.maxExecutionTimeInSeconds = 0;
+CLI usage: refexamples.maxExecutionTimeInSeconds = 0;
option name: minExecutionTimeInSeconds
description: algorithm will run at least specified seconds
values: class org.dllearner.core.options.IntegerConfigOption
default value: 0
-CLI usage: refinement.minExecutionTimeInSeconds = 0;
+CLI usage: refexamples.minExecutionTimeInSeconds = 0;
option name: guaranteeXgoodDescriptions
description: algorithm will run until X good (100%) concept descritpions are found
values: class org.dllearner.core.options.IntegerConfigOption
default value: 1
-CLI usage: refinement.guaranteeXgoodDescriptions = 1;
+CLI usage: refexamples.guaranteeXgoodDescriptions = 1;
+option name: maxClassDescriptionTests
+description: The maximum number of candidate hypothesis the algorithm is allowed to test (0 = no limit). The algorithm will stop afterwards. (The real number of tests can be slightly higher, because this criterion usually won't be checked after each single test.)
+values: class org.dllearner.core.options.IntegerConfigOption
+default value: 0
+CLI usage: refexamples.maxClassDescriptionTests = 0;
+
option name: logLevel
description: determines the logLevel for this component, can be {TRACE, DEBUG, INFO}
values: class org.dllearner.core.options.StringConfigOption
default value: DEBUG
-CLI usage: refinement.logLevel = DEBUG;
+CLI usage: refexamples.logLevel = DEBUG;
+option name: usePropernessChecks
+description: specifies whether to check for equivalence (i.e. discard equivalent refinements)
+values: class org.dllearner.core.options.BooleanConfigOption
+default value: false
+CLI usage: refexamples.usePropernessChecks = false;
+option name: maxPosOnlyExpansion
+description: specifies how often a node in the search tree of a posonly learning problem needs to be expanded before it is considered as solution candidate
+values: class org.dllearner.core.options.IntegerConfigOption
+default value: 4
+CLI usage: refexamples.maxPosOnlyExpansion = 4;
+
+option name: noisePercentage
+description: the (approximated) percentage of noise within the examples
+values: class org.dllearner.core.options.DoubleConfigOption min 0.0 max 100.0
+default value: 0.0
+CLI usage: refexamples.noisePercentage = 0.0;
+
+option name: terminateOnNoiseReached
+description: specifies whether to terminate when noise criterion is met
+values: class org.dllearner.core.options.BooleanConfigOption
+default value: true
+CLI usage: refexamples.terminateOnNoiseReached = true;
+
+option name: startClass
+description: the named class which should be used to start the algorithm (GUI: needs a widget for selecting a class)
+values: class org.dllearner.core.options.StringConfigOption
+default value: null
+CLI usage: refexamples.startClass = null;
+
+option name: forceRefinementLengthIncrease
+description: specifies whether nodes should be expanded until only longer refinements are reached
+values: class org.dllearner.core.options.BooleanConfigOption
+default value: null
+CLI usage: refexamples.forceRefinementLengthIncrease = null;
+
+option name: negativeWeight
+description: Used to penalise errors on negative examples different from those of positive examples (lower = less importance for negatives).
+values: class org.dllearner.core.options.DoubleConfigOption
+default value: 1.0
+CLI usage: refexamples.negativeWeight = 1.0;
+
+option name: startNodeBonus
+description: You can use this to give a heuristic bonus on the start node (= initially broader exploration of search space).
+values: class org.dllearner.core.options.DoubleConfigOption
+default value: 0.0
+CLI usage: refexamples.startNodeBonus = 0.0;
+
+option name: negationPenalty
+description: Penalty on negations (TODO: better explanation).
+values: class org.dllearner.core.options.IntegerConfigOption
+default value: 0
+CLI usage: refexamples.negationPenalty = 0;
+
+option name: expansionPenaltyFactor
+description: describes the reduction in heuristic score one is willing to accept for reducing the length of the concept by one
+values: class org.dllearner.core.options.DoubleConfigOption
+default value: 0.02
+CLI usage: refexamples.expansionPenaltyFactor = 0.02;
+
+
Modified: trunk/doc/manual/manual.tex
===================================================================
--- trunk/doc/manual/manual.tex 2009-02-17 15:28:41 UTC (rev 1607)
+++ trunk/doc/manual/manual.tex 2009-02-17 19:18:32 UTC (rev 1608)
@@ -54,7 +54,7 @@
In the most common scenario we consider, supervised learning in OWL/DLs, means that we have a background knowledge base in OWL/DLs. Additionally, we are given positive and negative examples. Each example is an individual in our knowledge base. The goal is to find an OWL \emph{class expression}\footnote{\owlce} such that all/many of the positive examples are \emph{instances} of this expression and none/few of the negative examples are instances of it. The primary purpose of learning is to find a class expression, which can classify unseen individuals (i.e.~not belonging to the examples) correctly. It is also important that the obtained class expression is easy to understand for a domain expert. We call these criteria \emph{accuracy} and \emph{readability}.
-As an example, consider the problem to find out whether a chemical compound can cause cancer\footnote{see \carc{} for a more detailed description}. In this case, the background knowledge contains information about chemical compounds in general and certain compounds we are interested in. The positive examples are those compounds causing cancer, whereas the negative examples are those compounds not causing cancer. The prediction for the examples is likely to have been obtained from experiments and long-term research trials in this case. Of course, all examples have to be present in the considered background knowledge. A learning algorithm can now derive a class expression from examples and background knowledge, e.g.~this class expression in natural language could be ``chemical compounds containing a phosphorus atom''. (Of course, in practice the expression will be more complex to obtain a reasonable accuracy.) Using this class expression, we can no classify unseen chemical compounds.
+As an example, consider the problem to find out whether a chemical compound can cause cancer\footnote{see \carc{} for a more detailed description}. In this case, the background knowledge contains information about chemical compounds in general and certain compounds we are interested in. The positive examples are those compounds causing cancer, whereas the negative examples are those compounds not causing cancer. The prediction for the examples is likely to have been obtained from experiments and long-term research trials in this case. Of course, all examples have to be present in the considered background knowledge. A learning algorithm can now derive a class expression from examples and background knowledge, e.g.~this class expression in natural language could be ``chemical compounds containing a phosphorus atom''. (Of course, in practice the expression will be more complex to obtain a reasonable accuracy.) Using this class expression, we can not classify unseen chemical compounds.
\section{Getting Started}
\label{sec:start}
@@ -119,22 +119,25 @@
\section{DL-Learner Components}
\label{sec:components}
-In this part, we describe concrete components currently implemented in DL-Learner. Each of the subsections contains a list of components of type specified in the subsection heading. Note that this constitute a full description, i.e.~we omit some components and many configuration options. The purpose of the manual is to obtain a general understanding of the implemented components. A full list of components, which is generated automatically from the source code, can be found in \verb|doc/configOptions.txt| including the default values for all options and their usage in conf files. \todo{ConfigDocumentationGenerator seems partially broken, because it prints things like CLI usage: null = owlAPIReasoner;}
+In this part, we describe concrete components currently implemented in DL-Learner. Each of the subsections contains a list of components according to the type specified in the subsection heading. Note that this does not constitute a full description, i.e.~we omit some components and many configuration options. The purpose of the manual is to obtain a general understanding of the implemented components. A full list of components, which is generated automatically from the source code, can be found in \verb|doc/configOptions.txt| including the default values for all options and their usage in conf files. \todo{ConfigDocumentationGenerator seems partially broken, because it prints things like CLI usage: null = owlAPIReasoner;}
\subsection{Knowledge Sources}
Knowledge sources have a URI and can be included in conf files using \verb|import("$url");|, e.g.~\verb|import("ontology.owl")|. Depending on the file ending, DL-Learner will guess the correct type of knowledge source. If you want to overwrite this, you can use a second parameter with value \verb|OWL|, \verb|KB|, or \verb|SPARQL|, e.g.~\verb|import("ontology.owl","OWL")|.
\begin{description}
- \item[OWL File] DL-Learner supports OWL files in different formats, e.g. RDF/XML or N-Triples. If there is a standard OWL format, you want to use, but is not supported by DL-Learner please let us know.
+ \item[OWL File] DL-Learner supports OWL files in different formats, e.g. RDF/XML or N-Triples. If there is a standard OWL format, you want to use, but is not supported by DL-Learner please let us know. Note: As we use the OWL API for parsing, all formats supported by it are possible\footnote{ for a list see \url{http://owlapi.sourceforge.net/}}.
\item[KB File] KB files are an internal non-standardised knowledge representation format, which corresponds to description logic syntax except that the special symbols have been replaced by ASCII strings, e.g.~\verb|AND| instead of $\sqcap$. You can find several KB files in the examples folder.
- \item[SPARQL Endpoint] DL-Learner allows to use SPARQL endpoints as background knowledge source, which enables the incorporation of very large knowledge bases, e.g. DBpedia\cite{2008_dbpedia}, in DL-Learner. This works by using a set of start instances, which usually correspond to the examples in a learning problem, and then retrieving knowledge about these instances via SPARQL queries. The obtained knowledge base fragment can be converted to OWL and consumed by a reasoner later since it is now sufficiently to be processed in reasonable time. Please see \cite{2009_ijswis} for details. Some options of the SPARQL component are:
+ \item[SPARQL Endpoint] DL-Learner allows to use SPARQL endpoints as background knowledge source, which enables the incorporation of very large knowledge bases, e.g. DBpedia\cite{2008_dbpedia}, in DL-Learner. This works by using a set of start instances, which usually correspond to the examples in a learning problem, and then retrieving knowledge about these instances via SPARQL queries. The obtained knowledge base fragment can be converted to OWL and consumed by a reasoner later since it is now sufficiently small to be processed in reasonable time. Please see \cite{2009_ijswis} for details, the parameters for the extraction algorithm are similar (different names only) to the ones here. Some options of the SPARQL component are:
\begin{itemize}
\item instances: Set of individuals to use for starting the knowledge fragment extraction. Example use in conf file: \begin{verbatim}sparql.instances = {"http://dbpedia.org/resource/Matt_Stone",
"http://dbpedia.org/resource/Sarah_Silverman"};\end{verbatim}
- \item recursionDepth: Maximum distance of an extracted individual from a start individual. This influences the size of the extracted fragment and depends on the maximum property depth you want the learned class expression to have. Example use in conf file: \verb|sparql.recursionDepth = 3;|.
+ \item recursionDepth: Maximum distance of an extracted individual from a start individual. This influences the size of the extracted fragment and depends on the maximum property depth you want the learned class expression to have. Example use in conf file: \verb|sparql.recursionDepth = 1;|.
+\item pecifies whether the extracted ontology is written to a file or not. The OWL file is written to the cache dir. The output is produced by OWL API, which does not check URI validity, but rather writes invalid XML.
+\verb|sparql.saveExtractedFragment = true;|
\end{itemize}
Many further options allow to modify the extracted fragment on the fly or fine-tune the extraction process.
+The extraction can be started separately by running and modifying \verb|org.dllearner.test.SparqlExtractionTest|. The collected ontology will be saved in the root dir.
\end{description}
\subsection{Reasoner Components}
@@ -147,12 +150,13 @@
\item reasonerType: Selects the desired reasoner. By default, Pellet is used. Usage: \verb|owlAPIReasoner.reasonerType = fact;|. Note that FaCT++ is written in C++ and we currently ship the 32 bit version of the JNI layer. This may change to 64 bit in the future.
\end{itemize}
\item[DIG] DIG 1.1\footnote{\dig} is an interface to description logic reasoners and supported by a large variety of reasoners including Pellet, FaCT++, KAON2, and Racer Pro. The major drawback is that the current version DIG 1.1 is not aligned with the OWL specification and therefore lacks several features, which are crucial to the more recent learning algorithms in DL-Learner. If you still want to use the DIG interface, you have to download a DIG capable reasoner and start the DIG server there. DL-Learner communicates with the reasoner using the XML based protocol over HTTP.
- \item[Fast Instance Checker] Instance checks, i.e.~testing whether an individual is instance of a class, is the major reasoner task in many learning algorithms. This reasoner is a self-developed for the DL-Learner project. It remedies some problems related to Machine Learning and the Open World Assumption in OWL and therefore is not correct w.r.t.~OWL semantics. (See \cite{cheng00} Section 4 for an explanation.) Furthermore, it provides an improved performance for instance checks by precomputing some inferences. The fast instance checker is build on top of Pellet and the default reasoner component in DL-Learner.
+ \item[Fast Instance Checker] Instance checks, i.e.~testing whether an individual is instance of a class, is the major reasoner task in many learning algorithms. This reasoner is a self-development of the DL-Learner project. It remedies some problems related to Machine Learning and the Open World Assumption in OWL and therefore is not correct w.r.t.~OWL semantics. (See \cite{cheng00} Section 4 for an explanation.) Furthermore, it provides an improved performance for instance checks by precomputing some inferences. The fast instance checker is build on top of Pellet and the default reasoner component in DL-Learner.
\end{description}
\subsection{Learning Problems}
In the introduction Sections \ref{sec:whatis} and \ref{sec:start}, we described a specific learning problem where positive and negative examples are given. In practice different variations of similar problems occur. You can switch between the different problems using \verb|problem=$value;|, where \verb|$value| is one of \verb|posNegDefinitionLP|, \verb|posOnlyDefinitionLP|, \verb|classLearning|.
+Default is \verb|posNegDefinitionLP;|.
\begin{description}
\item[Positive a...
[truncated message content] |
