001/** 002 * Copyright (C) 2007 - 2016, Jens Lehmann 003 * 004 * This file is part of DL-Learner. 005 * 006 * DL-Learner is free software; you can redistribute it and/or modify 007 * it under the terms of the GNU General Public License as published by 008 * the Free Software Foundation; either version 3 of the License, or 009 * (at your option) any later version. 010 * 011 * DL-Learner is distributed in the hope that it will be useful, 012 * but WITHOUT ANY WARRANTY; without even the implied warranty of 013 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 014 * GNU General Public License for more details. 015 * 016 * You should have received a copy of the GNU General Public License 017 * along with this program. If not, see <http://www.gnu.org/licenses/>. 018 */ 019package org.dllearner.accuracymethods; 020 021import org.dllearner.core.AbstractReasonerComponent; 022import org.dllearner.core.ComponentAnn; 023import org.dllearner.core.Reasoner; 024import org.dllearner.core.config.ConfigOption; 025import org.dllearner.learningproblems.Heuristics; 026import org.semanticweb.owlapi.model.OWLClassExpression; 027import org.semanticweb.owlapi.model.OWLIndividual; 028 029import java.util.Collection; 030 031@ComponentAnn(name = "FMeasure Approximate", shortName = "approx.fmeasure", version = 0) 032public class AccMethodFMeasureApprox extends AccMethodFMeasure implements AccMethodTwoValuedApproximate, AccMethodWithBeta { 033 @ConfigOption(description = "The Approximate Delta", defaultValue = "0.05", required = false) 034 private double approxDelta = 0.05; 035 @ConfigOption(description = "reasoner component (configured by learning problem)") 036 private Reasoner reasoner; 037 038 public AccMethodFMeasureApprox(boolean init, AbstractReasonerComponent reasoner) { 039 this.setReasoner(reasoner); 040 if(init) { 041 init(); 042 } 043 } 044 045 public AccMethodFMeasureApprox() { 046 } 047 048 @Override 049 public double getAccApprox2(OWLClassExpression description, 050 Collection<OWLIndividual> positiveExamples, 051 Collection<OWLIndividual> negativeExamples, double noise) { 052 // System.out.println("Testing " + description); 053 054 // we abort when there are too many uncovered positives 055 int maxNotCovered = (int) Math.ceil(noise*positiveExamples.size()); 056 int instancesCovered = 0; 057 int instancesNotCovered = 0; 058 059 for(OWLIndividual ind : positiveExamples) { 060 if(reasoner.hasType(description, ind)) { 061 instancesCovered++; 062 } else { 063 instancesNotCovered ++; 064 if(instancesNotCovered > maxNotCovered) { 065 return -1; 066 } 067 } 068 } 069 070 double recall = instancesCovered/(double)positiveExamples.size(); 071 072 int testsPerformed = 0; 073 int instancesDescription = 0; 074 075 for(OWLIndividual ind : negativeExamples) { 076 077 if(reasoner.hasType(description, ind)) { 078 instancesDescription++; 079 } 080 testsPerformed++; 081 082 // check whether approximation is sufficiently accurate 083 double[] approx = Heuristics.getFScoreApproximation(instancesCovered, recall, (beta == 0 ? 1 : beta), negativeExamples.size(), testsPerformed, instancesDescription); 084 if(approx[1]<approxDelta) { 085 return approx[0]; 086 } 087 088 } 089 090 // standard computation (no approximation) 091 double precision = instancesCovered/(double)(instancesDescription+instancesCovered); 092 return Heuristics.getFScore(recall, precision, (beta == 0 ? 1 : beta)); 093 } 094 095 @Override 096 public double getApproxDelta() { 097 return approxDelta; 098 } 099 100 @Override 101 public void setApproxDelta(double approxDelta) { 102 this.approxDelta = approxDelta; 103 } 104 105 @Override 106 public void setReasoner(Reasoner reasoner) { 107 this.reasoner = reasoner; 108 } 109 110}