listIDTotal, listTreeTotal, listRelTotal, listArgTotal = lib.filterData(listIDTotal, listTreeTotal, listRelTotal, listArgTotal) lib.makePredicatenode(listIDTotal, listTreeTotal, listRelTotal, listArgTotal) print 'Getting Feature' # listLabel, listFeature, listCount = lib.getFeature(listIDAfterChunking, listTreeAfterChunking, listRelAfterChunking, listArgAfterChunking, listWordName, listCluster) listLabel, listFeature, listCount = lib.getFeatureAllNode(listIDTotal, listTreeTotal, listRelTotal, listArgTotal, listWordName, listCluster) lib.checkUnderscore(listIDTotal, listTreeTotal, listRelTotal, listArgTotal) listLabel = lib.getListLabelReduce(listLabel, listLabelOriginal) listLE, leLabel, listEncode = lib.getLabelEncoderParameter(listFeature, listLabel) print 'Separating Data' # groupInfo, groupListLabel, groupListFeature = lib.kFold(listID1Rel, listTree1Rel, listRel1Rel, listArg1Rel, listWordName, listCluster, foldNumber, listLabelOriginal) groupInfo, groupListLabel, groupListFeature, listOfListNumArgPerSen = lib.kFold(listIDTotal, listTreeTotal, listRelTotal, listArgTotal, listWordName, listCluster, foldNumber, listLabelOriginal) print 'Transforming Data' listOfListFeatureTrain, listOfListFeatureTest, listOfListLabelTrain, listOfListLabelTest, listOfListNumArg, listOfListPredicateType = lib.crossValidation(groupListLabel, groupListFeature, listLE, leLabel, foldNumber, groupInfo) print 'Running Algorithm' listPrecisionChunking, listRecallChunking, listF1ScoreChunking, listPrecisionClassify1Stage, listRecallClassify1Stage, listF1ScoreClassify1Stage, listPrecision1Stage, listRecall1Stage, listF1Score1Stage, listOfListLabelPredict, listOfListLabelILP, listDensityMatrix, listOfListVariable, listPrecisionArg, listRecallArg, listF1ScoreArg = lib.crossValidationSVM1Stage(listOfListFeatureTrain, listOfListFeatureTest, listOfListLabelTrain, listOfListLabelTest, listFeatureName, foldNumber, listOfListNumArg, listEncode, listOfListNumArgPerSen, listOfListPredicateType) # listPrecisionChunking, listRecallChunking, listF1ScoreChunking, listPrecisionIden, listRecallIden, listF1ScoreIden, listPrecisionClass, listRecallClass, listF1ScoreClass, listPrecision2Stage, listRecall2Stage, listF1Score2Stage = lib.crossValidationSVM2Stage(listOfListFeatureTrain, listOfListFeatureTest, listOfListLabelTrain, listOfListLabelTest, listFeatureIdenName, listFeatureClassName, foldNumber, listOfListNumArg, listEncode) print 'Output' F1Score = lib.output1Stage(listPrecisionChunking, listRecallChunking, listF1ScoreChunking, listPrecisionClassify1Stage, listRecallClassify1Stage, listF1ScoreClassify1Stage, listPrecision1Stage, listRecall1Stage, listF1Score1Stage, listPrecisionArg, listRecallArg, listF1ScoreArg) # F1Score = lib.output2Stage(listPrecisionChunking, listRecallChunking, listF1ScoreChunking, listPrecisionIden, listRecallIden, listF1ScoreIden, listPrecisionClass, listRecallClass, listF1ScoreClass, listPrecision2Stage, listRecall2Stage, listF1Score2Stage) endTime = datetime.now() print "Running time: " print (endTime - startTime)
listIDTotal, listTreeTotal, listRelTotal, listArgTotal = lib.mergeData(listID1Rel, listTree1Rel, listRel1Rel, listArg1Rel, listIDExtractFromMutliRel, listTreeExtractFromMutliRel, listRelExtractFromMutliRel, listArgExtractFromMutliRel) # listIDAfterChunking, listTreeAfterChunking, listRelAfterChunking, listArgAfterChunking = lib.chunking(listID1Rel, listTree1Rel, listRel1Rel, listArg1Rel) #listIDAfterChunking, listTreeAfterChunking, listRelAfterChunking, listArgAfterChunking = lib.chunking(listIDTotal, listTreeTotal, listRelTotal, listArgTotal) listIDTotal, listTreeTotal, listRelTotal, listArgTotal = lib.filterData(listIDTotal, listTreeTotal, listRelTotal, listArgTotal) listIDAfterChunking, listTreeAfterChunking, listRelAfterChunking, listArgAfterChunking = lib.chunking(listIDTotal, listTreeTotal, listRelTotal, listArgTotal) # lib.omitUnderscore(listIDAfterChunking, listTreeAfterChunking, listRelAfterChunking, listArgAfterChunking) print 'Getting Feature' listLabel, listFeature = lib.getFeature(listIDAfterChunking, listTreeAfterChunking, listRelAfterChunking, listArgAfterChunking, listWordName, listCluster) listLabel = lib.getListLabelReduce(listLabel, listLabelOriginal) listLE, leLabel, listEncode = lib.getLabelEncoderParameter(listFeature, listLabel) print 'Separating Data' # groupInfo, groupListLabel, groupListFeature = lib.kFold(listID1Rel, listTree1Rel, listRel1Rel, listArg1Rel, listWordName, listCluster, foldNumber, listLabelOriginal) groupInfo, groupListLabel, groupListFeature = lib.kFold(listIDTotal, listTreeTotal, listRelTotal, listArgTotal, listWordName, listCluster, foldNumber, listLabelOriginal) print 'Transforming Data' listOfListFeatureTrain, listOfListFeatureTest, listOfListLabelTrain, listOfListLabelTest, listOfListNumArg = lib.crossValidation(groupListLabel, groupListFeature, listLE, leLabel, foldNumber, groupInfo) print 'Running Algorithm' listPrecisionChunking, listRecallChunking, listF1ScoreChunking, listPrecisionClassify1Stage, listRecallClassify1Stage, listF1ScoreClassify1Stage, listPrecision1Stage, listRecall1Stage, listF1Score1Stage = lib.crossValidationSVM1Stage(listOfListFeatureTrain, listOfListFeatureTest, listOfListLabelTrain, listOfListLabelTest, listFeatureName, foldNumber, listOfListNumArg, listEncode) #listPrecisionChunking, listRecallChunking, listF1ScoreChunking, listPrecisionIden, listRecallIden, listF1ScoreIden, listPrecisionClass, listRecallClass, listF1ScoreClass, listPrecision2Stage, listRecall2Stage, listF1Score2Stage = lib.crossValidationMaxEnt2Stage(listOfListFeatureTrain, listOfListFeatureTest, listOfListLabelTrain, listOfListLabelTest, listFeatureIdenName, listFeatureClassName, foldNumber, listOfListNumArg, listEncode) print 'Output' F1Score = lib.output1Stage(listPrecisionChunking, listRecallChunking, listF1ScoreChunking, listPrecisionClassify1Stage, listRecallClassify1Stage, listF1ScoreClassify1Stage, listPrecision1Stage, listRecall1Stage, listF1Score1Stage) #F1Score = lib.output2Stage(listPrecisionChunking, listRecallChunking, listF1ScoreChunking, listPrecisionIden, listRecallIden, listF1ScoreIden, listPrecisionClass, listRecallClass, listF1ScoreClass, listPrecision2Stage, listRecall2Stage, listF1Score2Stage) endTime = datetime.now() print "Running time: " print (endTime - startTime)