def __init__(self, style=None, classSet=None, featureSet=None, gazetteerFileName=None, skiplist=None):
if classSet == None:
classSet = IdSet(1)
if featureSet == None:
featureSet = IdSet()
ExampleBuilder.__init__(self, classSet, featureSet)
assert( classSet.getId("neg") == 1 )
#gazetteerFileName="/usr/share/biotext/GeniaChallenge/SharedTaskTriggerTest/gazetteer-train"
if gazetteerFileName!=None:
self.gazetteer=Gazetteer.loadGztr(gazetteerFileName)
print >> sys.stderr, "Loaded gazetteer from",gazetteerFileName
else:
print >> sys.stderr, "No gazetteer loaded"
self.gazetteer=None
self._setDefaultParameters(["rel_features", "wordnet", "bb_features", "giuliano",
"epi_merge_negated", "limit_merged_types", "genia_task1",
"names", "build_for_nameless", "skip_for_nameless",
"pos_only", "all_tokens", "pos_pairs", "linear_ngrams",
"phospho", "drugbank_features", "ddi13_features", "metamap",
"only_types", "ontobiotope_features", "bb_spans", "w2v",
"no_context"])
self.styles = self.getParameters(style)
# if "selftrain_group" in self.styles:
# self.selfTrainGroups = set()
# if "selftrain_group-1" in self.styles:
# self.selfTrainGroups.add("-1")
# if "selftrain_group0" in self.styles:
# self.selfTrainGroups.add("0")
# if "selftrain_group1" in self.styles:
# self.selfTrainGroups.add("1")
# if "selftrain_group2" in self.styles:
# self.selfTrainGroups.add("2")
# if "selftrain_group3" in self.styles:
# self.selfTrainGroups.add("3")
# print >> sys.stderr, "Self-train-groups:", self.selfTrainGroups
self.skiplist = set()
if skiplist != None:
f = open(skiplist, "rt")
for line in f.readlines():
self.skiplist.add(line.strip())
f.close()
if self.styles["rel_features"]:
self.relFeatureBuilder = RELFeatureBuilder(featureSet)
if self.styles["wordnet"]:
self.wordNetFeatureBuilder = WordNetFeatureBuilder(featureSet)
if self.styles["bb_features"]:
self.bacteriaTokens = PhraseTriggerExampleBuilder.getBacteriaTokens()
#self.bacteriaTokens = PhraseTriggerExampleBuilder.getBacteriaTokens(PhraseTriggerExampleBuilder.getBacteriaNames())
if self.styles["giuliano"]:
self.giulianoFeatureBuilder = GiulianoFeatureBuilder(featureSet)
if self.styles["drugbank_features"]:
self.drugFeatureBuilder = DrugFeatureBuilder(featureSet)
if self.styles["ontobiotope_features"]:
self.ontobiotopeFeatureBuilder = OntoBiotopeFeatureBuilder(self.featureSet)
if self.styles["w2v"]:
self.wordVectorFeatureBuilder = WordVectorFeatureBuilder(featureSet)
def __init__(self, style=None, length=None, types=[], featureSet=None, classSet=None):
if featureSet == None:
featureSet = IdSet()
if classSet == None:
classSet = IdSet(1)
else:
classSet = classSet
assert( classSet.getId("neg") == 1 or (len(classSet.Ids)== 2 and classSet.getId("neg") == -1) )
ExampleBuilder.__init__(self, classSet=classSet, featureSet=featureSet)
self.styles = self.getParameters(style, [
"typed", "directed", "headsOnly", "graph_kernel", "noAnnType", "noMasking", "maxFeatures",
"genia_limits", "epi_limits", "id_limits", "rel_limits", "bb_limits", "bi_limits", "co_limits",
"genia_task1", "ontology", "nodalida", "bacteria_renaming", "trigger_features", "rel_features",
"ddi_features", "evex", "giuliano", "random", "themeOnly", "causeOnly", "no_path", "entities",
"skip_extra_triggers", "headsOnly", "graph_kernel", "trigger_features", "no_task", "no_dependency",
"disable_entity_features", "disable_terminus_features", "disable_single_element_features",
"disable_ngram_features", "disable_path_edge_features", "no_linear", "subset", "binary", "pos_only",
"entity_type"
])
if style == None: # no parameters given
style["typed"] = style["directed"] = style["headsOnly"] = True
# self.styles = style
# if "selftrain_group" in self.styles:
# self.selfTrainGroups = set()
# if "selftrain_group-1" in self.styles:
# self.selfTrainGroups.add("-1")
# if "selftrain_group0" in self.styles:
# self.selfTrainGroups.add("0")
# if "selftrain_group1" in self.styles:
# self.selfTrainGroups.add("1")
# if "selftrain_group2" in self.styles:
# self.selfTrainGroups.add("2")
# if "selftrain_group3" in self.styles:
# self.selfTrainGroups.add("3")
# print >> sys.stderr, "Self-train-groups:", self.selfTrainGroups
self.multiEdgeFeatureBuilder = MultiEdgeFeatureBuilder(self.featureSet)
# NOTE Temporarily re-enabling predicted range
#self.multiEdgeFeatureBuilder.definePredictedValueRange([], None)
if self.styles["graph_kernel"]:
from FeatureBuilders.GraphKernelFeatureBuilder import GraphKernelFeatureBuilder
self.graphKernelFeatureBuilder = GraphKernelFeatureBuilder(self.featureSet)
if self.styles["noAnnType"]:
self.multiEdgeFeatureBuilder.noAnnType = True
if self.styles["noMasking"]:
self.multiEdgeFeatureBuilder.maskNamedEntities = False
if self.styles["maxFeatures"]:
self.multiEdgeFeatureBuilder.maximum = True
if self.styles["genia_task1"]:
self.multiEdgeFeatureBuilder.filterAnnTypes.add("Entity")
self.tokenFeatureBuilder = TokenFeatureBuilder(self.featureSet)
if self.styles["ontology"]:
self.multiEdgeFeatureBuilder.ontologyFeatureBuilder = BioInferOntologyFeatureBuilder(self.featureSet)
if self.styles["nodalida"]:
self.nodalidaFeatureBuilder = NodalidaFeatureBuilder(self.featureSet)
if self.styles["bacteria_renaming"]:
self.bacteriaRenamingFeatureBuilder = BacteriaRenamingFeatureBuilder(self.featureSet)
if self.styles["trigger_features"]:
self.triggerFeatureBuilder = TriggerFeatureBuilder(self.featureSet)
self.triggerFeatureBuilder.useNonNameEntities = True
if self.styles["genia_task1"]:
self.triggerFeatureBuilder.filterAnnTypes.add("Entity")
#self.bioinferOntologies = OntologyUtils.loadOntologies(OntologyUtils.g_bioInferFileName)
if self.styles["rel_features"]:
self.relFeatureBuilder = RELFeatureBuilder(featureSet)
if self.styles["ddi_features"]:
self.drugFeatureBuilder = DrugFeatureBuilder(featureSet)
if self.styles["evex"]:
self.evexFeatureBuilder = EVEXFeatureBuilder(featureSet)
if self.styles["giuliano"]:
self.giulianoFeatureBuilder = GiulianoFeatureBuilder(featureSet)
self.pathLengths = length
assert(self.pathLengths == None)
self.types = types
if self.styles["random"]:
from FeatureBuilders.RandomFeatureBuilder import RandomFeatureBuilder
self.randomFeatureBuilder = RandomFeatureBuilder(self.featureSet)
class MultiEdgeExampleBuilder(ExampleBuilder):
"""
This example builder makes edge examples, i.e. examples describing
the event arguments.
"""
def __init__(self, style=None, length=None, types=[], featureSet=None, classSet=None):
if featureSet == None:
featureSet = IdSet()
if classSet == None:
classSet = IdSet(1)
else:
classSet = classSet
assert( classSet.getId("neg") == 1 or (len(classSet.Ids)== 2 and classSet.getId("neg") == -1) )
ExampleBuilder.__init__(self, classSet=classSet, featureSet=featureSet)
self.styles = self.getParameters(style, [
"typed", "directed", "headsOnly", "graph_kernel", "noAnnType", "noMasking", "maxFeatures",
"genia_limits", "epi_limits", "id_limits", "rel_limits", "bb_limits", "bi_limits", "co_limits",
"genia_task1", "ontology", "nodalida", "bacteria_renaming", "trigger_features", "rel_features",
"ddi_features", "evex", "giuliano", "random", "themeOnly", "causeOnly", "no_path", "entities",
"skip_extra_triggers", "headsOnly", "graph_kernel", "trigger_features", "no_task", "no_dependency",
"disable_entity_features", "disable_terminus_features", "disable_single_element_features",
"disable_ngram_features", "disable_path_edge_features", "no_linear", "subset", "binary", "pos_only",
"entity_type"
])
if style == None: # no parameters given
style["typed"] = style["directed"] = style["headsOnly"] = True
# self.styles = style
# if "selftrain_group" in self.styles:
# self.selfTrainGroups = set()
# if "selftrain_group-1" in self.styles:
# self.selfTrainGroups.add("-1")
# if "selftrain_group0" in self.styles:
# self.selfTrainGroups.add("0")
# if "selftrain_group1" in self.styles:
# self.selfTrainGroups.add("1")
# if "selftrain_group2" in self.styles:
# self.selfTrainGroups.add("2")
# if "selftrain_group3" in self.styles:
# self.selfTrainGroups.add("3")
# print >> sys.stderr, "Self-train-groups:", self.selfTrainGroups
self.multiEdgeFeatureBuilder = MultiEdgeFeatureBuilder(self.featureSet)
# NOTE Temporarily re-enabling predicted range
#self.multiEdgeFeatureBuilder.definePredictedValueRange([], None)
if self.styles["graph_kernel"]:
from FeatureBuilders.GraphKernelFeatureBuilder import GraphKernelFeatureBuilder
self.graphKernelFeatureBuilder = GraphKernelFeatureBuilder(self.featureSet)
if self.styles["noAnnType"]:
self.multiEdgeFeatureBuilder.noAnnType = True
if self.styles["noMasking"]:
self.multiEdgeFeatureBuilder.maskNamedEntities = False
if self.styles["maxFeatures"]:
self.multiEdgeFeatureBuilder.maximum = True
if self.styles["genia_task1"]:
self.multiEdgeFeatureBuilder.filterAnnTypes.add("Entity")
self.tokenFeatureBuilder = TokenFeatureBuilder(self.featureSet)
if self.styles["ontology"]:
self.multiEdgeFeatureBuilder.ontologyFeatureBuilder = BioInferOntologyFeatureBuilder(self.featureSet)
if self.styles["nodalida"]:
self.nodalidaFeatureBuilder = NodalidaFeatureBuilder(self.featureSet)
if self.styles["bacteria_renaming"]:
self.bacteriaRenamingFeatureBuilder = BacteriaRenamingFeatureBuilder(self.featureSet)
if self.styles["trigger_features"]:
self.triggerFeatureBuilder = TriggerFeatureBuilder(self.featureSet)
self.triggerFeatureBuilder.useNonNameEntities = True
if self.styles["genia_task1"]:
self.triggerFeatureBuilder.filterAnnTypes.add("Entity")
#self.bioinferOntologies = OntologyUtils.loadOntologies(OntologyUtils.g_bioInferFileName)
if self.styles["rel_features"]:
self.relFeatureBuilder = RELFeatureBuilder(featureSet)
if self.styles["ddi_features"]:
self.drugFeatureBuilder = DrugFeatureBuilder(featureSet)
if self.styles["evex"]:
self.evexFeatureBuilder = EVEXFeatureBuilder(featureSet)
if self.styles["giuliano"]:
self.giulianoFeatureBuilder = GiulianoFeatureBuilder(featureSet)
self.pathLengths = length
assert(self.pathLengths == None)
self.types = types
if self.styles["random"]:
from FeatureBuilders.RandomFeatureBuilder import RandomFeatureBuilder
self.randomFeatureBuilder = RandomFeatureBuilder(self.featureSet)
def definePredictedValueRange(self, sentences, elementName):
self.multiEdgeFeatureBuilder.definePredictedValueRange(sentences, elementName)
def getPredictedValueRange(self):
return self.multiEdgeFeatureBuilder.predictedRange
def filterEdgesByType(self, edges, typesToInclude):
if len(typesToInclude) == 0:
return edges
edgesToKeep = []
for edge in edges:
if edge.get("type") in typesToInclude:
edgesToKeep.append(edge)
return edgesToKeep
def getCategoryNameFromTokens(self, sentenceGraph, t1, t2, directed=True):
"""
Example class. Multiple overlapping edges create a merged type.
"""
types = set()
# if sentenceGraph.interactionGraph.has_edge(t1, t2):
# intEdges = sentenceGraph.interactionGraph.get_edge_data(t1, t2, default={})
# # NOTE: Only works if keys are ordered integers
# for i in range(len(intEdges)):
# types.add(intEdges[i]["element"].get("type"))
# if (not directed) and sentenceGraph.interactionGraph.has_edge(t2, t1):
# intEdges = sentenceGraph.interactionGraph.get_edge(t2, t1, default={})
# # NOTE: Only works if keys are ordered integers
# for i in range(len(intEdges)):
# types.add(intEdges[i]["element"].get("type"))
intEdges = sentenceGraph.interactionGraph.getEdges(t1, t2)
if (not directed):
intEdges = intEdges + sentenceGraph.interactionGraph.getEdges(t2, t1)
for intEdge in intEdges:
types.add(intEdge[2].get("type"))
types = list(types)
types.sort()
categoryName = ""
for name in types:
if categoryName != "":
categoryName += "---"
categoryName += name
if categoryName != "":
return categoryName
else:
return "neg"
def getCategoryName(self, sentenceGraph, e1, e2, directed=True, duplicateEntities=None):
"""
Example class. Multiple overlapping edges create a merged type.
"""
# interactions = []
# e1s = [e1]
# if duplicateEntities != None and e1 in duplicateEntities:
# e1s += duplicateEntities[e1]
# e2s = [e2]
# if duplicateEntities != None and e2 in duplicateEntities:
# e2s += duplicateEntities[e2]
# for entity1 in e1s:
# for entity2 in e2s:
# interactions = interactions + sentenceGraph.getInteractions(entity1, entity2)
# if not directed:
# interactions = interactions + sentenceGraph.getInteractions(entity2, entity1)
interactions = sentenceGraph.getInteractions(e1, e2, True)
#print interactions
types = set()
for interaction in interactions:
types.add(interaction[2].get("type"))
types = list(types)
types.sort()
categoryName = ""
for name in types:
if self.styles["causeOnly"] and name != "Cause":
continue
if self.styles["themeOnly"] and name != "Theme":
continue
if categoryName != "":
categoryName += "---"
categoryName += name
if categoryName != "":
return categoryName
else:
return "neg"
def isPotentialRELInteraction(self, e1, e2):
if e1.get("type") == "Protein" and e2.get("type") == "Entity":
return True
else:
return False
def isPotentialBBInteraction(self, e1, e2, sentenceGraph):
#if e1.get("type") == "Bacterium" and e2.get("type") in ["Host", "HostPart", "Geographical", "Environmental", "Food", "Medical", "Soil", "Water"]:
# Note: "Environment" type is misspelled as "Environmental" in the BB-task documentation
if e1.get("type") == "Bacterium" and e2.get("type") in ["Host", "HostPart", "Geographical", "Environment", "Food", "Medical", "Soil", "Water"]:
return True
elif e1.get("type") == "Host" and e2.get("type") == "HostPart":
return True
else:
return False
def getBISuperType(self, eType):
if eType in ["GeneProduct", "Protein", "ProteinFamily", "PolymeraseComplex"]:
return "ProteinEntity"
elif eType in ["Gene", "GeneFamily", "GeneComplex", "Regulon", "Site", "Promoter"]:
return "GeneEntity"
else:
return None
def isPotentialBIInteraction(self, e1, e2, sentenceGraph, stats):
e1Type = e1.get("type")
e1SuperType = self.getBISuperType(e1Type)
e2Type = e2.get("type")
e2SuperType = self.getBISuperType(e2Type)
tag = "(" + e1Type + "/" + e2Type + ")"
if e1Type == "Regulon":
if e2SuperType in ["GeneEntity", "ProteinEntity"]:
return True
if e1SuperType == "ProteinEntity":
if e2Type in ["Site", "Promoter", "Gene", "GeneComplex"]:
return True
if e1Type in ["Action", "Transcription", "Expression"]:
return True
if e1Type == "Site":
if e2SuperType == "GeneEntity":
return True
if e1Type == "Promoter":
if e2SuperType in ["GeneEntity", "ProteinEntity"]:
return True
if e1SuperType in ["GeneEntity", "ProteinEntity"]:
if e2SuperType in ["GeneEntity", "ProteinEntity"]:
return True
stats.filter("bi_limits") #+tag)
return False
def isPotentialEPIInteraction(self, e1, e2, sentenceGraph):
if e1.get("type") != "Catalysis":
if e1.get("type") in ["Protein", "Entity"]:
return False
elif e2.get("type") in ["Protein", "Entity"]:
return True
else:
return False
else: # Catalysis
if e2.get("type") != "Entity":
return True
else:
return False
assert False, (e1.get("type"), e2.get("type"))
def isPotentialIDInteraction(self, e1, e2, sentenceGraph):
e1Type = e1.get("type")
e2Type = e2.get("type")
e1IsCore = e1Type in ["Protein", "Regulon-operon", "Two-component-system", "Chemical", "Organism"]
e2IsCore = e2Type in ["Protein", "Regulon-operon", "Two-component-system", "Chemical", "Organism"]
if e1IsCore:
return False
elif e1Type in ["Gene_expression", "Transcription"]:
if e2Type in ["Protein", "Regulon-operon"]:
return True
else:
return False
elif e1Type in ["Protein_catabolism", "Phosphorylation"]:
if e2Type == "Protein":
return True
else:
return False
elif e1Type == "Localization":
if e2IsCore or e2Type == "Entity":
return True
else:
return False
elif e1Type in ["Binding", "Process"]:
if e2IsCore:
return True
else:
return False
elif "egulation" in e1Type:
if e2Type != "Entity":
return True
else:
return False
elif e1Type == "Entity":
if e2IsCore:
return True
else:
return False
assert False, (e1Type, e2Type)
def isPotentialCOInteraction(self, e1, e2, sentenceGraph):
if e1.get("type") == "Exp" and e2.get("type") == "Exp":
anaphoraTok = sentenceGraph.entityHeadTokenByEntity[e1]
antecedentTok = sentenceGraph.entityHeadTokenByEntity[e2]
antecedentTokenFound = False
for token in sentenceGraph.tokens:
if token == antecedentTok:
antecedentTokenFound = True
if token == anaphoraTok: # if, not elif, to take into accoutn cases where e1Tok == e2Tok
if antecedentTokenFound:
return True
else:
return False
assert False
elif e1.get("type") == "Exp" and e2.get("type") == "Protein":
return True
else:
return False
def isPotentialGeniaInteraction(self, e1, e2):
e1Type = e1.get("type")
e2Type = e2.get("type")
if e1Type == "Protein":
return False
elif e1Type in ["Entity", "Gene_expression", "Transcription", "Protein_catabolism", "Phosphorylation", "Binding"]:
if e2Type == "Protein":
return True
else:
return False
elif e1Type == "Localization":
if e2Type in ["Protein", "Entity"]:
return True
else:
return False
elif "egulation" in e1Type:
if e2Type != "Entity":
return True
else:
return False
assert False, (e1Type, e2Type)
def getGoldCategoryName(self, goldGraph, entityToGold, e1, e2, directed=True):
if len(entityToGold[e1]) > 0 and len(entityToGold[e2]) > 0:
return self.getCategoryName(goldGraph, entityToGold[e1][0], entityToGold[e2][0], directed=directed)
else:
return "neg"
def buildExamplesFromGraph(self, sentenceGraph, outfile, goldGraph = None):
"""
Build examples for a single sentence. Returns a list of examples.
See Core/ExampleUtils for example format.
"""
#examples = []
exampleIndex = 0
if self.styles["trigger_features"]:
self.triggerFeatureBuilder.initSentence(sentenceGraph)
if self.styles["evex"]:
self.evexFeatureBuilder.initSentence(sentenceGraph)
# Filter entities, if needed
#mergedIds = None
#duplicateEntities = None
#entities = sentenceGraph.entities
#entities, mergedIds, duplicateEntities = self.mergeEntities(sentenceGraph, False) # "no_duplicates" in self.styles)
sentenceGraph.mergeInteractionGraph(True)
entities = sentenceGraph.mergedEntities
entityToDuplicates = sentenceGraph.mergedEntityToDuplicates
self.exampleStats.addValue("Duplicate entities skipped", len(sentenceGraph.entities) - len(entities))
# Connect to optional gold graph
if goldGraph != None:
entityToGold = EvaluateInteractionXML.mapEntities(entities, goldGraph.entities)
paths = None
if not self.styles["no_path"]:
##undirected = sentenceGraph.getUndirectedDependencyGraph()
#undirected = self.nxMultiDiGraphToUndirected(sentenceGraph.dependencyGraph)
###undirected = sentenceGraph.dependencyGraph.to_undirected()
####undirected = NX10.MultiGraph(sentenceGraph.dependencyGraph) This didn't work
undirected = sentenceGraph.dependencyGraph.toUndirected()
#paths = NX10.all_pairs_shortest_path(undirected, cutoff=999)
paths = undirected
#for edge in sentenceGraph.dependencyGraph.edges:
# assert edge[2] != None
#for edge in undirected.edges:
# assert edge[2] != None
#if sentenceGraph.sentenceElement.get("id") == "GENIA.d70.s5":
# print [(x[0].get("id"), x[1].get("id"), x[2].get("id")) for x in sentenceGraph.dependencyGraph.edges]
# Generate examples based on interactions between entities or interactions between tokens
if self.styles["entities"]:
loopRange = len(entities)
else:
loopRange = len(sentenceGraph.tokens)
for i in range(loopRange-1):
for j in range(i+1,loopRange):
eI = None
eJ = None
if self.styles["entities"]:
eI = entities[i]
eJ = entities[j]
tI = sentenceGraph.entityHeadTokenByEntity[eI]
tJ = sentenceGraph.entityHeadTokenByEntity[eJ]
#if "no_ne_interactions" in self.styles and eI.get("isName") == "True" and eJ.get("isName") == "True":
# continue
if eI.get("type") == "neg" or eJ.get("type") == "neg":
continue
if self.styles["skip_extra_triggers"]:
if eI.get("source") != None or eJ.get("source") != None:
continue
else:
tI = sentenceGraph.tokens[i]
tJ = sentenceGraph.tokens[j]
# only consider paths between entities (NOTE! entities, not only named entities)
if self.styles["headsOnly"]:
if (len(sentenceGraph.tokenIsEntityHead[tI]) == 0) or (len(sentenceGraph.tokenIsEntityHead[tJ]) == 0):
continue
if self.styles["directed"]:
# define forward
if self.styles["entities"]:
categoryName = self.getCategoryName(sentenceGraph, eI, eJ, True)
if goldGraph != None:
categoryName = self.getGoldCategoryName(goldGraph, entityToGold, eI, eJ, True)
else:
categoryName = self.getCategoryNameFromTokens(sentenceGraph, tI, tJ, True)
# make forward
self.exampleStats.beginExample(categoryName)
makeExample = True
if self.styles["genia_limits"] and not self.isPotentialGeniaInteraction(eI, eJ):
makeExample = False
self.exampleStats.filter("genia_limits")
if self.styles["genia_task1"] and (eI.get("type") == "Entity" or eJ.get("type") == "Entity"):
makeExample = False
self.exampleStats.filter("genia_task1")
if self.styles["rel_limits"] and not self.isPotentialRELInteraction(eI, eJ):
makeExample = False
self.exampleStats.filter("rel_limits")
if self.styles["co_limits"] and not self.isPotentialCOInteraction(eI, eJ, sentenceGraph):
makeExample = False
self.exampleStats.filter("co_limits")
if self.styles["bb_limits"] and not self.isPotentialBBInteraction(eI, eJ, sentenceGraph):
makeExample = False
self.exampleStats.filter("bb_limits")
if categoryName != "neg":
self.exampleStats.filter("bb_limits(" + categoryName + ":" + eI.get("type") + "/" + eJ.get("type") + ")")
if self.styles["bi_limits"] and not self.isPotentialBIInteraction(eI, eJ, sentenceGraph, self.exampleStats):
makeExample = False
#self.exampleStats.filter("bi_limits")
if self.styles["epi_limits"] and not self.isPotentialEPIInteraction(eI, eJ, sentenceGraph):
makeExample = False
self.exampleStats.filter("epi_limits")
if self.styles["id_limits"] and not self.isPotentialIDInteraction(eI, eJ, sentenceGraph):
makeExample = False
self.exampleStats.filter("id_limits")
# if self.styles["selftrain_limits"] and (eI.get("selftrain") == "False" or eJ.get("selftrain") == "False"):
# makeExample = False
# self.exampleStats.filter("selftrain_limits")
# if self.styles["selftrain_group"] and (eI.get("selftraingroup") not in self.selfTrainGroups or eJ.get("selftraingroup") not in self.selfTrainGroups):
# makeExample = False
# self.exampleStats.filter("selftrain_group")
if self.styles["pos_only"] and categoryName == "neg":
makeExample = False
self.exampleStats.filter("pos_only")
if makeExample:
#examples.append( self.buildExample(tI, tJ, paths, sentenceGraph, categoryName, exampleIndex, eI, eJ) )
ExampleUtils.appendExamples([self.buildExample(tI, tJ, paths, sentenceGraph, categoryName, exampleIndex, eI, eJ)], outfile)
exampleIndex += 1
self.exampleStats.endExample()
# define reverse
if self.styles["entities"]:
categoryName = self.getCategoryName(sentenceGraph, eJ, eI, True)
if goldGraph != None:
categoryName = self.getGoldCategoryName(goldGraph, entityToGold, eJ, eI, True)
else:
categoryName = self.getCategoryNameFromTokens(sentenceGraph, tJ, tI, True)
# make reverse
self.exampleStats.beginExample(categoryName)
makeExample = True
if self.styles["genia_limits"] and not self.isPotentialGeniaInteraction(eJ, eI):
makeExample = False
self.exampleStats.filter("genia_limits")
if self.styles["genia_task1"] and (eI.get("type") == "Entity" or eJ.get("type") == "Entity"):
makeExample = False
self.exampleStats.filter("genia_task1")
if self.styles["rel_limits"] and not self.isPotentialRELInteraction(eJ, eI):
makeExample = False
self.exampleStats.filter("rel_limits")
if self.styles["co_limits"] and not self.isPotentialCOInteraction(eJ, eI, sentenceGraph):
makeExample = False
self.exampleStats.filter("co_limits")
if self.styles["bb_limits"] and not self.isPotentialBBInteraction(eJ, eI, sentenceGraph):
makeExample = False
self.exampleStats.filter("bb_limits")
if categoryName != "neg":
self.exampleStats.filter("bb_limits(" + categoryName + ":" + eJ.get("type") + "/" + eI.get("type") + ")")
if self.styles["bi_limits"] and not self.isPotentialBIInteraction(eJ, eI, sentenceGraph, self.exampleStats):
makeExample = False
#self.exampleStats.filter("bi_limits")
if self.styles["epi_limits"] and not self.isPotentialEPIInteraction(eJ, eI, sentenceGraph):
makeExample = False
self.exampleStats.filter("epi_limits")
if self.styles["id_limits"] and not self.isPotentialIDInteraction(eJ, eI, sentenceGraph):
makeExample = False
self.exampleStats.filter("id_limits")
# if self.styles["selftrain_limits"] and (eI.get("selftrain") == "False" or eJ.get("selftrain") == "False"):
# makeExample = False
# self.exampleStats.filter("selftrain_limits")
# if self.styles["selftrain_group"] and (eI.get("selftraingroup") not in self.selfTrainGroups or eJ.get("selftraingroup") not in self.selfTrainGroups):
# makeExample = False
# self.exampleStats.filter("selftrain_group")
if self.styles["pos_only"] and categoryName == "neg":
makeExample = False
self.exampleStats.filter("pos_only")
if makeExample:
#examples.append( self.buildExample(tJ, tI, paths, sentenceGraph, categoryName, exampleIndex, eJ, eI) )
ExampleUtils.appendExamples([self.buildExample(tJ, tI, paths, sentenceGraph, categoryName, exampleIndex, eJ, eI)], outfile)
exampleIndex += 1
self.exampleStats.endExample()
else:
if self.styles["entities"]:
categoryName = self.getCategoryName(sentenceGraph, eI, eJ, False)
else:
categoryName = self.getCategoryNameFromTokens(sentenceGraph, tI, tJ, False)
self.exampleStats.beginExample(categoryName)
forwardExample = self.buildExample(tI, tJ, paths, sentenceGraph, categoryName, exampleIndex, eI, eJ)
if not self.styles["graph_kernel"]:
reverseExample = self.buildExample(tJ, tI, paths, sentenceGraph, categoryName, exampleIndex, eJ, eI)
forwardExample[2].update(reverseExample[2])
#examples.append(forwardExample)
ExampleUtils.appendExamples([forwardExample], outfile)
exampleIndex += 1
self.exampleStats.endExample()
#return examples
return exampleIndex
def buildExample(self, token1, token2, paths, sentenceGraph, categoryName, exampleIndex, entity1=None, entity2=None):
"""
Build a single directed example for the potential edge between token1 and token2
"""
# dummy return for speed testing
#return (sentenceGraph.getSentenceId()+".x"+str(exampleIndex),1,{},{})
# define features
features = {}
if True: #token1 != token2 and paths.has_key(token1) and paths[token1].has_key(token2):
#if token1 != token2 and paths.has_key(token1) and paths[token1].has_key(token2):
# path = paths[token1][token2]
#else:
# path = [token1, token2]
if not self.styles["no_path"]:
# directedPath reduces performance by 0.01 pp
#directedPath = sentenceGraph.dependencyGraph.getPaths(token1, token2)
#if len(directedPath) == 0:
# directedPath = sentenceGraph.dependencyGraph.getPaths(token2, token1)
# for dp in directedPath:
# dp.reverse()
#if len(directedPath) == 0:
# path = paths.getPaths(token1, token2)
#else:
# path = directedPath
path = paths.getPaths(token1, token2)
if len(path) > 0:
#if len(path) > 1:
# print len(path)
path = path[0]
pathExists = True
else:
path = [token1, token2]
pathExists = False
else:
path = [token1, token2]
pathExists = False
#print token1.get("id"), token2.get("id")
assert(self.pathLengths == None)
if self.pathLengths == None or len(path)-1 in self.pathLengths:
# if not "no_ontology" in self.styles:
# self.ontologyFeatureBuilder.setFeatureVector(features)
# self.ontologyFeatureBuilder.buildOntologyFeaturesForPath(sentenceGraph, path)
# self.ontologyFeatureBuilder.setFeatureVector(None)
if self.styles["trigger_features"]: # F 85.52 -> 85.55
self.triggerFeatureBuilder.setFeatureVector(features)
self.triggerFeatureBuilder.tag = "trg1_"
self.triggerFeatureBuilder.buildFeatures(token1)
self.triggerFeatureBuilder.tag = "trg2_"
self.triggerFeatureBuilder.buildFeatures(token2)
self.triggerFeatureBuilder.setFeatureVector(None)
# REL features
if self.styles["rel_features"] and not self.styles["no_task"]:
self.relFeatureBuilder.setFeatureVector(features)
self.relFeatureBuilder.tag = "rel1_"
self.relFeatureBuilder.buildAllFeatures(sentenceGraph.tokens, sentenceGraph.tokens.index(token1))
self.relFeatureBuilder.tag = "rel2_"
self.relFeatureBuilder.buildAllFeatures(sentenceGraph.tokens, sentenceGraph.tokens.index(token2))
self.relFeatureBuilder.setFeatureVector(None)
if self.styles["bacteria_renaming"] and not self.styles["no_task"]:
self.bacteriaRenamingFeatureBuilder.setFeatureVector(features)
self.bacteriaRenamingFeatureBuilder.buildPairFeatures(entity1, entity2)
#self.bacteriaRenamingFeatureBuilder.buildSubstringFeatures(entity1, entity2) # decreases perf. 74.76 -> 72.41
self.bacteriaRenamingFeatureBuilder.setFeatureVector(None)
if self.styles["co_limits"] and not self.styles["no_task"]:
e1Offset = Range.charOffsetToSingleTuple(entity1.get("charOffset"))
e2Offset = Range.charOffsetToSingleTuple(entity2.get("charOffset"))
if Range.contains(e1Offset, e2Offset):
features[self.featureSet.getId("e1_contains_e2")] = 1
if entity2.get("isName") == "True":
features[self.featureSet.getId("e1_contains_e2name")] = 1
if Range.contains(e2Offset, e1Offset):
features[self.featureSet.getId("e2_contains_e1")] = 1
if entity1.get("isName") == "True":
features[self.featureSet.getId("e2_contains_e1name")] = 1
if self.styles["ddi_features"]:
self.drugFeatureBuilder.setFeatureVector(features)
self.drugFeatureBuilder.tag = "ddi_"
self.drugFeatureBuilder.buildPairFeatures(entity1, entity2)
if self.styles["ddi_mtmx"]:
self.drugFeatureBuilder.buildMTMXFeatures(entity1, entity2)
self.drugFeatureBuilder.setFeatureVector(None)
#if "graph_kernel" in self.styles or not "no_dependency" in self.styles:
# #print "Getting edges"
# if token1 != token2 and pathExists:
# #print "g1"
# edges = self.multiEdgeFeatureBuilder.getEdges(sentenceGraph.dependencyGraph, path)
# #print "g2"
# else:
# edges = None
if self.styles["graph_kernel"]:
self.graphKernelFeatureBuilder.setFeatureVector(features, entity1, entity2)
self.graphKernelFeatureBuilder.buildGraphKernelFeatures(sentenceGraph, path)
self.graphKernelFeatureBuilder.setFeatureVector(None)
if self.styles["entity_type"]:
features[self.featureSet.getId("e1_"+entity1.get("type"))] = 1
features[self.featureSet.getId("e2_"+entity2.get("type"))] = 1
features[self.featureSet.getId("distance_"+str(len(path)))] = 1
if not self.styles["no_dependency"]:
#print "Dep features"
self.multiEdgeFeatureBuilder.setFeatureVector(features, entity1, entity2)
#self.multiEdgeFeatureBuilder.buildStructureFeatures(sentenceGraph, paths) # remove for fast
if not self.styles["disable_entity_features"]:
self.multiEdgeFeatureBuilder.buildEntityFeatures(sentenceGraph)
self.multiEdgeFeatureBuilder.buildPathLengthFeatures(path)
if not self.styles["disable_terminus_features"]:
self.multiEdgeFeatureBuilder.buildTerminusTokenFeatures(path, sentenceGraph) # remove for fast
if not self.styles["disable_single_element_features"]:
self.multiEdgeFeatureBuilder.buildSingleElementFeatures(path, sentenceGraph)
if not self.styles["disable_ngram_features"]:
#print "NGrams"
self.multiEdgeFeatureBuilder.buildPathGrams(2, path, sentenceGraph) # remove for fast
self.multiEdgeFeatureBuilder.buildPathGrams(3, path, sentenceGraph) # remove for fast
self.multiEdgeFeatureBuilder.buildPathGrams(4, path, sentenceGraph) # remove for fast
#self.buildEdgeCombinations(path, edges, sentenceGraph, features) # remove for fast
#if edges != None:
# self.multiEdgeFeatureBuilder.buildTerminusFeatures(path[0], edges[0][1]+edges[1][0], "t1", sentenceGraph) # remove for fast
# self.multiEdgeFeatureBuilder.buildTerminusFeatures(path[-1], edges[len(path)-1][len(path)-2]+edges[len(path)-2][len(path)-1], "t2", sentenceGraph) # remove for fast
if not self.styles["disable_path_edge_features"]:
self.multiEdgeFeatureBuilder.buildPathEdgeFeatures(path, sentenceGraph)
self.multiEdgeFeatureBuilder.buildSentenceFeatures(sentenceGraph)
self.multiEdgeFeatureBuilder.setFeatureVector(None)
if self.styles["nodalida"]:
self.nodalidaFeatureBuilder.setFeatureVector(features, entity1, entity2)
shortestPaths = self.nodalidaFeatureBuilder.buildShortestPaths(sentenceGraph.dependencyGraph, path)
print shortestPaths
if len(shortestPaths) > 0:
self.nodalidaFeatureBuilder.buildNGrams(shortestPaths, sentenceGraph)
self.nodalidaFeatureBuilder.setFeatureVector(None)
if not self.styles["no_linear"]:
self.tokenFeatureBuilder.setFeatureVector(features)
for i in range(len(sentenceGraph.tokens)):
if sentenceGraph.tokens[i] == token1:
token1Index = i
if sentenceGraph.tokens[i] == token2:
token2Index = i
linearPreTag = "linfw_"
if token1Index > token2Index:
token1Index, token2Index = token2Index, token1Index
linearPreTag = "linrv_"
self.tokenFeatureBuilder.buildLinearOrderFeatures(token1Index, sentenceGraph, 2, 2, preTag="linTok1")
self.tokenFeatureBuilder.buildLinearOrderFeatures(token2Index, sentenceGraph, 2, 2, preTag="linTok2")
# Before, middle, after
# self.tokenFeatureBuilder.buildTokenGrams(0, token1Index-1, sentenceGraph, "bf")
# self.tokenFeatureBuilder.buildTokenGrams(token1Index+1, token2Index-1, sentenceGraph, "bw")
# self.tokenFeatureBuilder.buildTokenGrams(token2Index+1, len(sentenceGraph.tokens)-1, sentenceGraph, "af")
# before-middle, middle, middle-after
# self.tokenFeatureBuilder.buildTokenGrams(0, token2Index-1, sentenceGraph, linearPreTag+"bf", max=2)
# self.tokenFeatureBuilder.buildTokenGrams(token1Index+1, token2Index-1, sentenceGraph, linearPreTag+"bw", max=2)
# self.tokenFeatureBuilder.buildTokenGrams(token1Index+1, len(sentenceGraph.tokens)-1, sentenceGraph, linearPreTag+"af", max=2)
self.tokenFeatureBuilder.setFeatureVector(None)
if self.styles["random"]:
self.randomFeatureBuilder.setFeatureVector(features)
self.randomFeatureBuilder.buildRandomFeatures(100, 0.01)
self.randomFeatureBuilder.setFeatureVector(None)
if self.styles["genia_limits"] and not self.styles["no_task"]:
e1Type = entity1.get("type")
e2Type = entity2.get("type")
assert(entity1.get("isName") == "False")
if entity2.get("isName") == "True":
features[self.featureSet.getId("GENIA_target_protein")] = 1
else:
features[self.featureSet.getId("GENIA_nested_event")] = 1
if e1Type.find("egulation") != -1: # leave r out to avoid problems with capitalization
if entity2.get("isName") == "True":
features[self.featureSet.getId("GENIA_regulation_of_protein")] = 1
else:
features[self.featureSet.getId("GENIA_regulation_of_event")] = 1
if self.styles["bi_limits"]:
# Make features based on entity types
e1Type = entity1.get("type")
e2Type = entity2.get("type")
e1SuperType = str(self.getBISuperType(e1Type))
e2SuperType = str(self.getBISuperType(e2Type))
features[self.featureSet.getId("BI_e1_"+e1Type)] = 1
features[self.featureSet.getId("BI_e2_"+e2Type)] = 1
features[self.featureSet.getId("BI_e1sup_"+e1SuperType)] = 1
features[self.featureSet.getId("BI_e2sup_"+e2SuperType)] = 1
features[self.featureSet.getId("BI_e1e2_"+e1Type+"_"+e2Type)] = 1
features[self.featureSet.getId("BI_e1e2sup_"+e1SuperType+"_"+e2SuperType)] = 1
if self.styles["evex"]:
self.evexFeatureBuilder.setFeatureVector(features, entity1, entity2)
self.evexFeatureBuilder.buildEdgeFeatures(entity1, entity2, token1, token2, path, sentenceGraph)
self.evexFeatureBuilder.setFeatureVector(None)
if self.styles["giuliano"]:
self.giulianoFeatureBuilder.setFeatureVector(features, entity1, entity2)
self.giulianoFeatureBuilder.buildEdgeFeatures(entity1, entity2, token1, token2, path, sentenceGraph)
self.giulianoFeatureBuilder.setFeatureVector(None)
else:
features[self.featureSet.getId("always_negative")] = 1
if self.styles["subset"]:
features[self.featureSet.getId("out_of_scope")] = 1
else:
features[self.featureSet.getId("always_negative")] = 1
if self.styles["subset"]:
features[self.featureSet.getId("out_of_scope")] = 1
path = [token1, token2]
# define extra attributes
#if int(path[0].get("id").split("_")[-1]) < int(path[-1].get("id").split("_")[-1]):
if int(path[0].get("charOffset").split("-")[0]) < int(path[-1].get("charOffset").split("-")[0]):
#extra = {"xtype":"edge","type":"i","t1":path[0],"t2":path[-1]}
extra = {"xtype":"edge","type":"i","t1":path[0].get("id"),"t2":path[-1].get("id")}
extra["deprev"] = False
else:
#extra = {"xtype":"edge","type":"i","t1":path[-1],"t2":path[0]}
extra = {"xtype":"edge","type":"i","t1":path[-1].get("id"),"t2":path[0].get("id")}
extra["deprev"] = True
if entity1 != None:
#extra["e1"] = entity1
extra["e1"] = entity1.get("id")
if sentenceGraph.mergedEntityToDuplicates != None:
#extra["e1GoldIds"] = mergedEntityIds[entity1]
extra["e1DuplicateIds"] = ",".join([x.get("id") for x in sentenceGraph.mergedEntityToDuplicates[entity1]])
if entity2 != None:
#extra["e2"] = entity2
extra["e2"] = entity2.get("id")
if sentenceGraph.mergedEntityToDuplicates != None:
extra["e2DuplicateIds"] = ",".join([x.get("id") for x in sentenceGraph.mergedEntityToDuplicates[entity2]])
#extra["e2GoldIds"] = mergedEntityIds[entity2]
extra["categoryName"] = categoryName
if self.styles["bacteria_renaming"]:
if entity1.get("text") != None and entity1.get("text") != "":
extra["e1t"] = entity1.get("text").replace(" ", "---").replace(":","-COL-")
if entity2.get("text") != None and entity2.get("text") != "":
extra["e2t"] = entity2.get("text").replace(" ", "---").replace(":","-COL-")
sentenceOrigId = sentenceGraph.sentenceElement.get("origId")
if sentenceOrigId != None:
extra["SOID"] = sentenceOrigId
# make example
if self.styles["binary"]:
if categoryName != "neg":
category = 1
else:
category = -1
categoryName = "i"
else:
category = self.classSet.getId(categoryName)
# NOTE: temporarily disable for replicating 110310 experiment
#features[self.featureSet.getId("extra_constant")] = 1
return (sentenceGraph.getSentenceId()+".x"+str(exampleIndex),category,features,extra)
class EntityExampleBuilder(ExampleBuilder):
def __init__(self, style=None, classSet=None, featureSet=None, gazetteerFileName=None, skiplist=None):
if classSet == None:
classSet = IdSet(1)
if featureSet == None:
featureSet = IdSet()
ExampleBuilder.__init__(self, classSet, featureSet)
assert( classSet.getId("neg") == 1 )
#gazetteerFileName="/usr/share/biotext/GeniaChallenge/SharedTaskTriggerTest/gazetteer-train"
if gazetteerFileName!=None:
self.gazetteer=Gazetteer.loadGztr(gazetteerFileName)
print >> sys.stderr, "Loaded gazetteer from",gazetteerFileName
else:
print >> sys.stderr, "No gazetteer loaded"
self.gazetteer=None
self._setDefaultParameters(["rel_features", "wordnet", "bb_features", "giuliano",
"epi_merge_negated", "limit_merged_types", "genia_task1",
"names", "build_for_nameless", "skip_for_nameless",
"pos_only", "all_tokens", "pos_pairs", "linear_ngrams",
"phospho", "drugbank_features", "ddi13_features", "metamap"])
self.styles = self.getParameters(style)
# if "selftrain_group" in self.styles:
# self.selfTrainGroups = set()
# if "selftrain_group-1" in self.styles:
# self.selfTrainGroups.add("-1")
# if "selftrain_group0" in self.styles:
# self.selfTrainGroups.add("0")
# if "selftrain_group1" in self.styles:
# self.selfTrainGroups.add("1")
# if "selftrain_group2" in self.styles:
# self.selfTrainGroups.add("2")
# if "selftrain_group3" in self.styles:
# self.selfTrainGroups.add("3")
# print >> sys.stderr, "Self-train-groups:", self.selfTrainGroups
self.skiplist = set()
if skiplist != None:
f = open(skiplist, "rt")
for line in f.readlines():
self.skiplist.add(line.strip())
f.close()
if self.styles["rel_features"]:
self.relFeatureBuilder = RELFeatureBuilder(featureSet)
if self.styles["wordnet"]:
self.wordNetFeatureBuilder = WordNetFeatureBuilder(featureSet)
if self.styles["bb_features"]:
self.bacteriaTokens = PhraseTriggerExampleBuilder.getBacteriaTokens()
#self.bacteriaTokens = PhraseTriggerExampleBuilder.getBacteriaTokens(PhraseTriggerExampleBuilder.getBacteriaNames())
if self.styles["giuliano"]:
self.giulianoFeatureBuilder = GiulianoFeatureBuilder(featureSet)
if self.styles["drugbank_features"]:
self.drugFeatureBuilder = DrugFeatureBuilder(featureSet)
def getMergedEntityType(self, entities):
"""
If a single token belongs to multiple entities of different types,
a new, composite type is defined. This type is the alphabetically
ordered types of these entities joined with '---'.
"""
types = set()
entityIds = set()
for entity in entities:
if entity.get("given") == "True" and self.styles["all_tokens"]:
continue
if entity.get("type") == "Entity" and self.styles["genia_task1"]:
continue
if self.styles["epi_merge_negated"]:
types.add(Utils.InteractionXML.ResolveEPITriggerTypes.getEPIBaseType(entity.get("type")))
entityIds.add(entity.get("id"))
else:
types.add(entity.get("type"))
entityIds.add(entity.get("id"))
types = list(types)
types.sort()
typeString = ""
for type in types:
#if type == "Protein" and "all_tokens" in self.styles:
# continue
if typeString != "":
typeString += "---"
typeString += type
if typeString == "":
return "neg", None
idString = "/".join(sorted(list(entityIds)))
if self.styles["limit_merged_types"]:
if typeString.find("---") != -1:
if typeString == "Gene_expression---Positive_regulation":
return typeString, idString
else:
return typeString.split("---")[0], idString # ids partially incorrect
else:
return typeString, idString
return typeString, idString
def getMetaMapFeatures(self, token, sentenceGraph, features):
analyses = sentenceGraph.sentenceElement.find("analyses")
if analyses == None:
return
metamap = analyses.find("metamap")
if metamap == None:
return
tokenOffset = Range.charOffsetToSingleTuple(token.get("charOffset"))
skipAttr = set(["charOffset", "text"])
for phrase in metamap.findall("phrase"):
phraseOffset = Range.charOffsetToSingleTuple(phrase.get("charOffset"))
if Range.overlap(tokenOffset, phraseOffset):
attr = phrase.attrib
attrNames = sorted(attr.keys())
for attrName in attrNames:
if attrName in skipAttr:
continue
elif attrName == "score":
features["_metamap_score"] = 0.001 * abs(int(attr[attrName]))
else:
attrValues = attr[attrName].split(",")
for attrValue in attrValues:
features["_metamap_"+attrName+"_"+attrValue.replace(" ", "-")] = 1
def getTokenFeatures(self, token, sentenceGraph):
"""
Returns a list of features based on the attributes of a token.
These can be used to define more complex features.
"""
# These features are cached when this method is first called
# for a token.
if self.tokenFeatures.has_key(token):
return self.tokenFeatures[token], self.tokenFeatureWeights[token]
tokTxt=sentenceGraph.getTokenText(token)
features = {}
features["_txt_"+tokTxt]=1
features["_POS_"+token.get("POS")]=1
if sentenceGraph.tokenIsName[token] and not self.styles["names"]:
features["_given"]=1
for entity in sentenceGraph.tokenIsEntityHead[token]:
if entity.get("given") == "True":
features["_annType_"+entity.get("type")]=1
if self.styles["metamap"]:
self.getMetaMapFeatures(token, sentenceGraph, features)
# # Filip's gazetteer based features (can be used separately from exclude_gazetteer)
# if "gazetteer_features" in self.styles:
# tokTxtLower = tokTxt.lower()
# if "stem_gazetteer" in self.styles:
# tokTxtLower = PorterStemmer.stem(tokTxtLower)
# if self.gazetteer and tokTxtLower in self.gazetteer:
# for label,weight in self.gazetteer[tokTxtLower].items():
# features["_knownLabel_"+label]=weight # 1 performs slightly worse
## BANNER features
#if sentenceGraph.entityHintsByToken.has_key(token):
# features["BANNER-entity"] = 1
# Wordnet features
#if "wordnet" in self.styles:
# for wordNetFeature in self.wordNetFeatureBuilder.getTokenFeatures(tokTxt, token.get("POS")):
# features["_WN_"+wordNetFeature] = 1
self.tokenFeatures[token] = sorted(features.keys())
self.tokenFeatureWeights[token] = features
return self.tokenFeatures[token], self.tokenFeatureWeights[token]
def buildLinearOrderFeatures(self,sentenceGraph,index,tag,features):
"""
Linear features are built by marking token features with a tag
that defines their relative position in the linear order.
"""
tag = "linear_"+tag
tokenFeatures, tokenFeatureWeights = self.getTokenFeatures(sentenceGraph.tokens[index], sentenceGraph)
for tokenFeature in tokenFeatures:
features[self.featureSet.getId(tag+tokenFeature)] = tokenFeatureWeights[tokenFeature]
def buildLinearNGram(self, i, j, sentenceGraph, features):
ngram = "ngram"
for index in range(i, j+1):
ngram += "_" + sentenceGraph.getTokenText(sentenceGraph.tokens[index]).lower()
features[self.featureSet.getId(ngram)] = 1
def buildExamplesFromGraph(self, sentenceGraph, outfile, goldGraph=None, structureAnalyzer=None):
"""
Build one example for each token of the sentence
"""
if sentenceGraph.sentenceElement.get("origId") in self.skiplist:
print >> sys.stderr, "Skipping sentence", sentenceGraph.sentenceElement.get("origId")
return 0 #[]
#examples = []
exampleIndex = 0
self.tokenFeatures = {}
self.tokenFeatureWeights = {}
# determine (manually or automatically) the setting for whether sentences with no given entities should be skipped
buildForNameless = False
if structureAnalyzer and not structureAnalyzer.hasGroupClass("GIVEN", "ENTITY"): # no given entities points to no separate NER program being used
buildForNameless = True
if self.styles["build_for_nameless"]: # manually force the setting
buildForNameless = True
if self.styles["skip_for_nameless"]: # manually force the setting
buildForNameless = False
# determine whether sentences with no given entities should be skipped
namedEntityHeadTokens = []
if not self.styles["names"]:
namedEntityCount = 0
for entity in sentenceGraph.entities:
if entity.get("given") == "True": # known data which can be used for features
namedEntityCount += 1
namedEntityCountFeature = "nameCount_" + str(namedEntityCount)
# NOTE!!! This will change the number of examples and omit
# all triggers (positive and negative) from sentences which
# have no NE:s, possibly giving a too-optimistic performance
# value. Such sentences can still have triggers from intersentence
# interactions, but as such events cannot be recovered anyway,
# looking for these triggers would be pointless.
if namedEntityCount == 0 and not buildForNameless: # no names, no need for triggers
return 0 #[]
if self.styles["pos_pairs"]:
namedEntityHeadTokens = self.getNamedEntityHeadTokens(sentenceGraph)
else:
for key in sentenceGraph.tokenIsName.keys():
sentenceGraph.tokenIsName[key] = False
bagOfWords = {}
for token in sentenceGraph.tokens:
text = "bow_" + token.get("text")
if not bagOfWords.has_key(text):
bagOfWords[text] = 0
bagOfWords[text] += 1
if sentenceGraph.tokenIsName[token]:
text = "ne_" + text
if not bagOfWords.has_key(text):
bagOfWords[text] = 0
bagOfWords[text] += 1
bowFeatures = {}
for k in sorted(bagOfWords.keys()):
bowFeatures[self.featureSet.getId(k)] = bagOfWords[k]
self.inEdgesByToken = {}
self.outEdgesByToken = {}
self.edgeSetByToken = {}
for token in sentenceGraph.tokens:
#inEdges = sentenceGraph.dependencyGraph.in_edges(token, data=True)
#fixedInEdges = []
#for edge in inEdges:
# fixedInEdges.append( (edge[0], edge[1], edge[2]["element"]) )
#inEdges = fixedInEdges
inEdges = sentenceGraph.dependencyGraph.getInEdges(token)
#inEdges.sort(compareDependencyEdgesById)
self.inEdgesByToken[token] = inEdges
#outEdges = sentenceGraph.dependencyGraph.out_edges(token, data=True)
#fixedOutEdges = []
#for edge in outEdges:
# fixedOutEdges.append( (edge[0], edge[1], edge[2]["element"]) )
#outEdges = fixedOutEdges
outEdges = sentenceGraph.dependencyGraph.getOutEdges(token)
#outEdges.sort(compareDependencyEdgesById)
self.outEdgesByToken[token] = outEdges
self.edgeSetByToken[token] = set(inEdges + outEdges)
for i in range(len(sentenceGraph.tokens)):
token = sentenceGraph.tokens[i]
# CLASS
if len(sentenceGraph.tokenIsEntityHead[token]) > 0:
categoryName, entityIds = self.getMergedEntityType(sentenceGraph.tokenIsEntityHead[token])
else:
categoryName, entityIds = "neg", None
self.exampleStats.beginExample(categoryName)
# Recognize only non-named entities (i.e. interaction words)
if sentenceGraph.tokenIsName[token] and not self.styles["names"] and not self.styles["all_tokens"]:
self.exampleStats.filter("name")
self.exampleStats.endExample()
continue
# if "selftrain_limits" in self.styles:
# # any predicted entity not part of the self-training set causes example to be rejected
# filtered = False
# for entity in sentenceGraph.tokenIsEntityHead[token]:
# if entity.get("selftrain") == "False":
# self.exampleStats.filter("selftrain_limits")
# self.exampleStats.endExample()
# filtered = True
# break
# if filtered:
# continue
# if "selftrain_group" in self.styles:
# # any predicted entity not part of the self-training set causes example to be rejected
# filtered = False
# for entity in sentenceGraph.tokenIsEntityHead[token]:
# if entity.get("selftraingroup") not in self.selfTrainGroups:
# self.exampleStats.filter("selftrain_group")
# self.exampleStats.endExample()
# filtered = True
# break
# if filtered:
# continue
if self.styles["pos_only"] and categoryName == "neg":
self.exampleStats.filter("pos_only")
self.exampleStats.endExample()
continue
category = self.classSet.getId(categoryName)
if category == None:
self.exampleStats.filter("undefined_class")
self.exampleStats.endExample()
continue
tokenText = token.get("text").lower()
# if "stem_gazetteer" in self.styles:
# tokenText = PorterStemmer.stem(tokenText)
# if ("exclude_gazetteer" in self.styles) and self.gazetteer and tokenText not in self.gazetteer:
# features = {}
# features[self.featureSet.getId("exclude_gazetteer")] = 1
# extra = {"xtype":"token","t":token.get("id"),"excluded":"True"}
# if entityIds != None:
# extra["goldIds"] = entityIds
# #examples.append( (sentenceGraph.getSentenceId()+".x"+str(exampleIndex),category,features,extra) )
# ExampleUtils.appendExamples([(sentenceGraph.getSentenceId()+".x"+str(exampleIndex),category,features,extra)], outfile)
# exampleIndex += 1
# continue
# FEATURES
features = {}
if not self.styles["names"]:
features[self.featureSet.getId(namedEntityCountFeature)] = 1
#for k,v in bagOfWords.iteritems():
# features[self.featureSet.getId(k)] = v
# pre-calculate bow _features_
features.update(bowFeatures)
# for j in range(len(sentenceGraph.tokens)):
# text = "bow_" + sentenceGraph.tokens[j].get("text")
# if j < i:
# features[self.featureSet.getId("bf_" + text)] = 1
# elif j > i:
# features[self.featureSet.getId("af_" + text)] = 1
# Main features
text = token.get("text")
features[self.featureSet.getId("txt_"+text)] = 1
features[self.featureSet.getId("POS_"+token.get("POS"))] = 1
stem = PorterStemmer.stem(text)
features[self.featureSet.getId("stem_"+stem)] = 1
features[self.featureSet.getId("nonstem_"+text[len(stem):])] = 1
# Normalized versions of the string (if same as non-normalized, overlap without effect)
normalizedText = text.replace("-","").replace("/","").replace(",","").replace("\\","").replace(" ","").lower()
if normalizedText == "bound": # should be for all irregular verbs
normalizedText = "bind"
features[self.featureSet.getId("txt_"+normalizedText)] = 1
norStem = PorterStemmer.stem(normalizedText)
features[self.featureSet.getId("stem_"+norStem)] = 1
features[self.featureSet.getId("nonstem_"+normalizedText[len(norStem):])] = 1
## Subspan features
#textLower = text.lower()
#for i in range(1, len(textLower)):
# features[self.featureSet.getId("subspanbegin"+str(i)+"_"+textLower[0:i])] = 1
# features[self.featureSet.getId("subspanend"+str(i)+"_"+textLower[-i:])] = 1
# Substring features
for string in text.split("-"):
stringLower = string.lower()
features[self.featureSet.getId("substring_"+stringLower)] = 1
features[self.featureSet.getId("substringstem_"+PorterStemmer.stem(stringLower))] = 1
# Linear order features
for index in [-3,-2,-1,1,2,3]:
if i + index > 0 and i + index < len(sentenceGraph.tokens):
self.buildLinearOrderFeatures(sentenceGraph, i + index, str(index), features)
# Linear n-grams
if self.styles["linear_ngrams"]:
self.buildLinearNGram(max(0, i-1), i, sentenceGraph, features)
self.buildLinearNGram(max(0, i-2), i, sentenceGraph, features)
if self.styles["phospho"]:
if text.find("hospho") != -1:
features[self.featureSet.getId("phospho_found")] = 1
features[self.featureSet.getId("begin_"+text[0:2].lower())] = 1
features[self.featureSet.getId("begin_"+text[0:3].lower())] = 1
if self.styles["bb_features"]:
if text.lower() in self.bacteriaTokens:
features[self.featureSet.getId("lpsnBacToken")] = 1
# Content
if i > 0 and text[0].isalpha() and text[0].isupper():
features[self.featureSet.getId("upper_case_start")] = 1
for j in range(len(text)):
if j > 0 and text[j].isalpha() and text[j].isupper():
features[self.featureSet.getId("upper_case_middle")] = 1
# numbers and special characters
if text[j].isdigit():
features[self.featureSet.getId("has_digits")] = 1
if j > 0 and text[j-1] == "-":
features[self.featureSet.getId("has_hyphenated_digit")] = 1
elif text[j] == "-":
features[self.featureSet.getId("has_hyphen")] = 1
elif text[j] == "/":
features[self.featureSet.getId("has_fslash")] = 1
elif text[j] == "\\":
features[self.featureSet.getId("has_bslash")] = 1
# duplets
if j > 0:
features[self.featureSet.getId("dt_"+text[j-1:j+1].lower())] = 1
# triplets
if j > 1:
features[self.featureSet.getId("tt_"+text[j-2:j+1].lower())] = 1
# quadruplets (don't work, slight decrease (0.5 pp) on f-score
#if j > 2:
# features[self.featureSet.getId("qt_"+text[j-3:j+1].lower())] = 1
# Attached edges (Hanging in and out edges)
t1InEdges = self.inEdgesByToken[token]
for edge in t1InEdges:
edgeType = edge[2].get("type")
features[self.featureSet.getId("t1HIn_"+edgeType)] = 1
features[self.featureSet.getId("t1HIn_"+edge[0].get("POS"))] = 1
features[self.featureSet.getId("t1HIn_"+edgeType+"_"+edge[0].get("POS"))] = 1
tokenText = sentenceGraph.getTokenText(edge[0])
features[self.featureSet.getId("t1HIn_"+tokenText)] = 1
features[self.featureSet.getId("t1HIn_"+edgeType+"_"+tokenText)] = 1
tokenStem = PorterStemmer.stem(tokenText)
features[self.featureSet.getId("t1HIn_"+tokenStem)] = 1
features[self.featureSet.getId("t1HIn_"+edgeType+"_"+tokenStem)] = 1
features[self.featureSet.getId("t1HIn_"+norStem+"_"+edgeType+"_"+tokenStem)] = 1
t1OutEdges = self.outEdgesByToken[token]
for edge in t1OutEdges:
edgeType = edge[2].get("type")
features[self.featureSet.getId("t1HOut_"+edgeType)] = 1
features[self.featureSet.getId("t1HOut_"+edge[1].get("POS"))] = 1
features[self.featureSet.getId("t1HOut_"+edgeType+"_"+edge[1].get("POS"))] = 1
tokenText = sentenceGraph.getTokenText(edge[1])
features[self.featureSet.getId("t1HOut_"+tokenText)] = 1
features[self.featureSet.getId("t1HOut_"+edgeType+"_"+tokenText)] = 1
tokenStem = PorterStemmer.stem(tokenText)
features[self.featureSet.getId("t1HOut_"+tokenStem)] = 1
features[self.featureSet.getId("t1HOut_"+edgeType+"_"+tokenStem)] = 1
features[self.featureSet.getId("t1HOut_"+norStem+"_"+edgeType+"_"+tokenStem)] = 1
# REL features
if self.styles["rel_features"]:
self.relFeatureBuilder.setFeatureVector(features)
self.relFeatureBuilder.buildAllFeatures(sentenceGraph.tokens, i)
self.relFeatureBuilder.setFeatureVector(None)
# DDI13 features
if self.styles["ddi13_features"]:
for index in range(len(normalizedText)):
features[self.featureSet.getId("ddi13_fromstart" + str(index) + "_" + normalizedText[:index+1])] = 1
features[self.featureSet.getId("ddi13_fromend" + str(index) + "_" + normalizedText[index:])] = 1
if self.styles["drugbank_features"]:
self.drugFeatureBuilder.setFeatureVector(features)
self.drugFeatureBuilder.tag = "ddi_"
self.drugFeatureBuilder.buildDrugFeatures(token)
self.drugFeatureBuilder.setFeatureVector(None)
#self.wordNetFeatureBuilder.getTokenFeatures("show", "VBP")
#tokTxt = token.get("text")
#tokPOS = token.get("POS")
#wordNetFeatures = []
#wordNetFeatures = self.wordNetFeatureBuilder.getTokenFeatures(tokTxt, tokPOS)
#self.wordNetFeatureBuilder.getTokenFeatures(tokTxt, tokPOS)
if self.styles["wordnet"]:
tokTxt = token.get("text")
tokPOS = token.get("POS")
wordNetFeatures = self.wordNetFeatureBuilder.getTokenFeatures(tokTxt, tokPOS)
for wordNetFeature in wordNetFeatures:
#print wordNetFeature,
features[self.featureSet.getId("WN_"+wordNetFeature)] = 1
#print
if self.styles["giuliano"]:
self.giulianoFeatureBuilder.setFeatureVector(features)
self.giulianoFeatureBuilder.buildTriggerFeatures(token, sentenceGraph)
self.giulianoFeatureBuilder.setFeatureVector(None)
extra = {"xtype":"token","t":token.get("id")}
if self.styles["bb_features"]:
extra["trigex"] = "bb" # Request trigger extension in ExampleWriter
if self.styles["epi_merge_negated"]:
extra["unmergeneg"] = "epi" # Request trigger type unmerging
if entityIds != None:
extra["goldIds"] = entityIds # The entities to which this example corresponds
#examples.append( (sentenceGraph.getSentenceId()+".x"+str(exampleIndex),category,features,extra) )
# chains
self.buildChains(token, sentenceGraph, features)
if self.styles["pos_pairs"]:
self.buildPOSPairs(token, namedEntityHeadTokens, features)
example = (sentenceGraph.getSentenceId()+".x"+str(exampleIndex), category, features, extra)
ExampleUtils.appendExamples([example], outfile)
exampleIndex += 1
self.exampleStats.endExample()
#return examples
return exampleIndex
def buildChains(self,token,sentenceGraph,features,depthLeft=3,chain="",visited=None):
if depthLeft == 0:
return
strDepthLeft = "dist_" + str(depthLeft)
if visited == None:
visited = set()
inEdges = self.inEdgesByToken[token]
outEdges = self.outEdgesByToken[token]
edgeSet = visited.union(self.edgeSetByToken[token])
for edge in inEdges:
if not edge in visited:
edgeType = edge[2].get("type")
features[self.featureSet.getId("dep_"+strDepthLeft+edgeType)] = 1
nextToken = edge[0]
tokenFeatures, tokenWeights = self.getTokenFeatures(nextToken, sentenceGraph)
for tokenFeature in tokenFeatures:
features[self.featureSet.getId(strDepthLeft + tokenFeature)] = tokenWeights[tokenFeature]
# for entity in sentenceGraph.tokenIsEntityHead[nextToken]:
# if entity.get("given") == "True":
# features[self.featureSet.getId("name_dist_"+strDepthLeft)] = 1
# features[self.featureSet.getId("name_dist_"+strDepthLeft+entity.get("type"))] = 1
# features[self.featureSet.getId("POS_dist_"+strDepthLeft+nextToken.get("POS"))] = 1
# tokenText = sentenceGraph.getTokenText(nextToken)
# features[self.featureSet.getId("text_dist_"+strDepthLeft+tokenText)] = 1
if sentenceGraph.tokenIsName[nextToken] and not self.styles["names"]:
features[self.featureSet.getId("name_chain_dist_"+strDepthLeft+chain+"-frw_"+edgeType)] = 1
features[self.featureSet.getId("chain_dist_"+strDepthLeft+chain+"-frw_"+edgeType)] = 1
self.buildChains(nextToken,sentenceGraph,features,depthLeft-1,chain+"-frw_"+edgeType,edgeSet)
for edge in outEdges:
if not edge in visited:
edgeType = edge[2].get("type")
features[self.featureSet.getId("dep_dist_"+strDepthLeft+edgeType)] = 1
nextToken = edge[1]
tokenFeatures, tokenWeights = self.getTokenFeatures(nextToken, sentenceGraph)
for tokenFeature in tokenFeatures:
features[self.featureSet.getId(strDepthLeft + tokenFeature)] = tokenWeights[tokenFeature]
# for entity in sentenceGraph.tokenIsEntityHead[nextToken]:
# if entity.get("given") == "True":
# features[self.featureSet.getId("name_dist_"+strDepthLeft)] = 1
# features[self.featureSet.getId("name_dist_"+strDepthLeft+entity.get("type"))] = 1
# features[self.featureSet.getId("POS_dist_"+strDepthLeft+nextToken.get("POS"))] = 1
# tokenText = sentenceGraph.getTokenText(nextToken)
# features[self.featureSet.getId("text_dist_"+strDepthLeft+tokenText)] = 1
if sentenceGraph.tokenIsName[nextToken] and not self.styles["names"]:
features[self.featureSet.getId("name_chain_dist_"+strDepthLeft+chain+"-rev_"+edgeType)] = 1
features[self.featureSet.getId("chain_dist_"+strDepthLeft+chain+"-rev_"+edgeType)] = 1
self.buildChains(nextToken,sentenceGraph,features,depthLeft-1,chain+"-rev_"+edgeType,edgeSet)
def getNamedEntityHeadTokens(self, sentenceGraph):
headTokens = []
for entity in sentenceGraph.entities:
if entity.get("given") == "True": # known data which can be used for features
headTokens.append(sentenceGraph.entityHeadTokenByEntity[entity])
return headTokens
def buildPOSPairs(self, token, namedEntityHeadTokens, features):
tokenPOS = token.get("POS")
assert tokenPOS != None
for headToken in namedEntityHeadTokens:
headPOS = headToken.get("POS")
features[self.featureSet.getId("POS_pair_NE_"+tokenPOS+"-"+headPOS)] = 1
def __init__(self, style=None, types=[], featureSet=None, classSet=None):
if featureSet == None:
featureSet = IdSet()
if classSet == None:
classSet = IdSet(1)
else:
classSet = classSet
ExampleBuilder.__init__(self, classSet=classSet, featureSet=featureSet)
assert( classSet.getId("neg") == 1 or (len(classSet.Ids)== 2 and classSet.getId("neg") == -1) )
# Basic style = trigger_features:typed:directed:no_linear:entities:auto_limits:noMasking:maxFeatures
self._setDefaultParameters([
"directed", "undirected", "headsOnly", "graph_kernel", "noAnnType", "mask_nodes", "limit_features",
"no_auto_limits", "co_features", "genia_features", "bi_features", #"genia_limits", "epi_limits", "id_limits", "rel_limits", "bb_limits", "bi_limits", "co_limits",
"genia_task1", "ontology", "nodalida", "bacteria_renaming", "no_trigger_features", "rel_features",
"drugbank_features", "ddi_mtmx", "evex", "giuliano", "random", "themeOnly", "causeOnly", "no_path", "token_nodes",
"skip_extra_triggers", "headsOnly", "graph_kernel", "no_task", "no_dependency",
"disable_entity_features", "disable_terminus_features", "disable_single_element_features",
"disable_ngram_features", "disable_path_edge_features", "linear_features", "subset", "binary", "pos_only",
"entity_type", "filter_shortest_path", "maskTypeAsProtein", "keep_neg", "metamap",
"sdb_merge", "sdb_features", "ontobiotope_features", "no_self_loops", "full_entities",
"no_features", "wordnet", "wordvector", "se10t8_undirected", "filter_types", "doc_extra",
"entity_extra"])
self.styles = self.getParameters(style)
#if style == None: # no parameters given
# style["typed"] = style["directed"] = style["headsOnly"] = True
self.multiEdgeFeatureBuilder = MultiEdgeFeatureBuilder(self.featureSet, self.styles)
# NOTE Temporarily re-enabling predicted range
#self.multiEdgeFeatureBuilder.definePredictedValueRange([], None)
if self.styles["graph_kernel"]:
from FeatureBuilders.GraphKernelFeatureBuilder import GraphKernelFeatureBuilder
self.graphKernelFeatureBuilder = GraphKernelFeatureBuilder(self.featureSet)
if self.styles["noAnnType"]:
self.multiEdgeFeatureBuilder.noAnnType = True
if self.styles["mask_nodes"]:
self.multiEdgeFeatureBuilder.maskNamedEntities = True
else:
self.multiEdgeFeatureBuilder.maskNamedEntities = False
if not self.styles["limit_features"]:
self.multiEdgeFeatureBuilder.maximum = True
if self.styles["genia_task1"]:
self.multiEdgeFeatureBuilder.filterAnnTypes.add("Entity")
self.tokenFeatureBuilder = TokenFeatureBuilder(self.featureSet)
if self.styles["ontology"]:
self.multiEdgeFeatureBuilder.ontologyFeatureBuilder = BioInferOntologyFeatureBuilder(self.featureSet)
if self.styles["ontobiotope_features"]:
self.ontobiotopeFeatureBuilder = OntoBiotopeFeatureBuilder(self.featureSet)
if self.styles["nodalida"]:
self.nodalidaFeatureBuilder = NodalidaFeatureBuilder(self.featureSet)
if self.styles["bacteria_renaming"]:
self.bacteriaRenamingFeatureBuilder = BacteriaRenamingFeatureBuilder(self.featureSet)
if not self.styles["no_trigger_features"]:
self.triggerFeatureBuilder = TriggerFeatureBuilder(self.featureSet, self.styles)
self.triggerFeatureBuilder.useNonNameEntities = True
if self.styles["noAnnType"]:
self.triggerFeatureBuilder.noAnnType = True
if self.styles["genia_task1"]:
self.triggerFeatureBuilder.filterAnnTypes.add("Entity")
#self.bioinferOntologies = OntologyUtils.loadOntologies(OntologyUtils.g_bioInferFileName)
if self.styles["rel_features"]:
self.relFeatureBuilder = RELFeatureBuilder(featureSet)
if self.styles["drugbank_features"]:
self.drugFeatureBuilder = DrugFeatureBuilder(featureSet)
if self.styles["evex"]:
self.evexFeatureBuilder = EVEXFeatureBuilder(featureSet)
if self.styles["wordnet"]:
self.wordNetFeatureBuilder = WordNetFeatureBuilder(featureSet)
if self.styles["wordvector"]:
self.wordVectorFeatureBuilder = WordVectorFeatureBuilder(featureSet, self.styles)
if self.styles["giuliano"]:
self.giulianoFeatureBuilder = GiulianoFeatureBuilder(featureSet)
self.types = types
if self.styles["random"]:
from FeatureBuilders.RandomFeatureBuilder import RandomFeatureBuilder
self.randomFeatureBuilder = RandomFeatureBuilder(self.featureSet)
class EdgeExampleBuilder(ExampleBuilder):
"""
This example builder makes edge examples, i.e. examples describing
the event arguments.
"""
def __init__(self, style=None, types=[], featureSet=None, classSet=None):
if featureSet == None:
featureSet = IdSet()
if classSet == None:
classSet = IdSet(1)
else:
classSet = classSet
ExampleBuilder.__init__(self, classSet=classSet, featureSet=featureSet)
assert( classSet.getId("neg") == 1 or (len(classSet.Ids)== 2 and classSet.getId("neg") == -1) )
# Basic style = trigger_features:typed:directed:no_linear:entities:auto_limits:noMasking:maxFeatures
self._setDefaultParameters([
"directed", "undirected", "headsOnly", "graph_kernel", "noAnnType", "mask_nodes", "limit_features",
"no_auto_limits", "co_features", "genia_features", "bi_features", #"genia_limits", "epi_limits", "id_limits", "rel_limits", "bb_limits", "bi_limits", "co_limits",
"genia_task1", "ontology", "nodalida", "bacteria_renaming", "no_trigger_features", "rel_features",
"drugbank_features", "ddi_mtmx", "evex", "giuliano", "random", "themeOnly", "causeOnly", "no_path", "token_nodes",
"skip_extra_triggers", "headsOnly", "graph_kernel", "no_task", "no_dependency",
"disable_entity_features", "disable_terminus_features", "disable_single_element_features",
"disable_ngram_features", "disable_path_edge_features", "linear_features", "subset", "binary", "pos_only",
"entity_type", "filter_shortest_path", "maskTypeAsProtein", "keep_neg", "metamap",
"sdb_merge", "sdb_features", "ontobiotope_features", "no_self_loops", "full_entities",
"no_features", "wordnet", "wordvector", "se10t8_undirected", "filter_types", "doc_extra",
"entity_extra"])
self.styles = self.getParameters(style)
#if style == None: # no parameters given
# style["typed"] = style["directed"] = style["headsOnly"] = True
self.multiEdgeFeatureBuilder = MultiEdgeFeatureBuilder(self.featureSet, self.styles)
# NOTE Temporarily re-enabling predicted range
#self.multiEdgeFeatureBuilder.definePredictedValueRange([], None)
if self.styles["graph_kernel"]:
from FeatureBuilders.GraphKernelFeatureBuilder import GraphKernelFeatureBuilder
self.graphKernelFeatureBuilder = GraphKernelFeatureBuilder(self.featureSet)
if self.styles["noAnnType"]:
self.multiEdgeFeatureBuilder.noAnnType = True
if self.styles["mask_nodes"]:
self.multiEdgeFeatureBuilder.maskNamedEntities = True
else:
self.multiEdgeFeatureBuilder.maskNamedEntities = False
if not self.styles["limit_features"]:
self.multiEdgeFeatureBuilder.maximum = True
if self.styles["genia_task1"]:
self.multiEdgeFeatureBuilder.filterAnnTypes.add("Entity")
self.tokenFeatureBuilder = TokenFeatureBuilder(self.featureSet)
if self.styles["ontology"]:
self.multiEdgeFeatureBuilder.ontologyFeatureBuilder = BioInferOntologyFeatureBuilder(self.featureSet)
if self.styles["ontobiotope_features"]:
self.ontobiotopeFeatureBuilder = OntoBiotopeFeatureBuilder(self.featureSet)
if self.styles["nodalida"]:
self.nodalidaFeatureBuilder = NodalidaFeatureBuilder(self.featureSet)
if self.styles["bacteria_renaming"]:
self.bacteriaRenamingFeatureBuilder = BacteriaRenamingFeatureBuilder(self.featureSet)
if not self.styles["no_trigger_features"]:
self.triggerFeatureBuilder = TriggerFeatureBuilder(self.featureSet, self.styles)
self.triggerFeatureBuilder.useNonNameEntities = True
if self.styles["noAnnType"]:
self.triggerFeatureBuilder.noAnnType = True
if self.styles["genia_task1"]:
self.triggerFeatureBuilder.filterAnnTypes.add("Entity")
#self.bioinferOntologies = OntologyUtils.loadOntologies(OntologyUtils.g_bioInferFileName)
if self.styles["rel_features"]:
self.relFeatureBuilder = RELFeatureBuilder(featureSet)
if self.styles["drugbank_features"]:
self.drugFeatureBuilder = DrugFeatureBuilder(featureSet)
if self.styles["evex"]:
self.evexFeatureBuilder = EVEXFeatureBuilder(featureSet)
if self.styles["wordnet"]:
self.wordNetFeatureBuilder = WordNetFeatureBuilder(featureSet)
if self.styles["wordvector"]:
self.wordVectorFeatureBuilder = WordVectorFeatureBuilder(featureSet, self.styles)
if self.styles["giuliano"]:
self.giulianoFeatureBuilder = GiulianoFeatureBuilder(featureSet)
self.types = types
if self.styles["random"]:
from FeatureBuilders.RandomFeatureBuilder import RandomFeatureBuilder
self.randomFeatureBuilder = RandomFeatureBuilder(self.featureSet)
def definePredictedValueRange(self, sentences, elementName):
self.multiEdgeFeatureBuilder.definePredictedValueRange(sentences, elementName)
def getPredictedValueRange(self):
return self.multiEdgeFeatureBuilder.predictedRange
def filterEdgesByType(self, edges, typesToInclude):
if len(typesToInclude) == 0:
return edges
edgesToKeep = []
for edge in edges:
if edge.get("type") in typesToInclude:
edgesToKeep.append(edge)
return edgesToKeep
def getCategoryNameFromTokens(self, sentenceGraph, t1, t2, directed=True):
"""
Example class. Multiple overlapping edges create a merged type.
"""
types = set()
intEdges = sentenceGraph.interactionGraph.getEdges(t1, t2)
if not directed:
intEdges = intEdges + sentenceGraph.interactionGraph.getEdges(t2, t1)
for intEdge in intEdges:
types.add(intEdge[2].get("type"))
types = list(types)
types.sort()
categoryName = ""
for name in types:
if categoryName != "":
categoryName += "---"
categoryName += name
if categoryName != "":
return categoryName
else:
return "neg"
def getCategoryName(self, sentenceGraph, e1, e2, directed=True):
"""
Example class. Multiple overlapping edges create a merged type.
"""
interactions = sentenceGraph.getInteractions(e1, e2, True)
if not directed and not self.styles["se10t8_undirected"]:
interactions = interactions + sentenceGraph.getInteractions(e2, e1, True)
types = set()
for interaction in interactions:
types.add(interaction[2].get("type"))
types = list(types)
types.sort()
categoryName = ""
for name in types:
if self.styles["causeOnly"] and name != "Cause":
continue
if self.styles["themeOnly"] and name != "Theme":
continue
if categoryName != "":
categoryName += "---"
if self.styles["sdb_merge"]:
name = self.mergeForSeeDev(name, self.structureAnalyzer)
categoryName += name
if categoryName != "":
return categoryName
else:
return "neg"
def getBISuperType(self, eType):
if eType in ["GeneProduct", "Protein", "ProteinFamily", "PolymeraseComplex"]:
return "ProteinEntity"
elif eType in ["Gene", "GeneFamily", "GeneComplex", "Regulon", "Site", "Promoter"]:
return "GeneEntity"
else:
return None
def getSeeDevSuperTypes(self, eType):
if eType in ("Gene", "Gene_Family", "Box", "Promoter"):
return ("DNA", "Molecule")
elif eType == "RNA":
return ("RNA", "DNA_Product", "Molecule")
elif eType in ("Protein", "Protein_Family", "Protein_Complex", "Protein_Domain"):
return ("Amino_acid_sequence", "DNA_Product", "Molecule")
elif eType == "Hormone":
return ("Molecule",)
elif eType in ("Regulatory_Network", "Pathway"):
return ("Dynamic_process",)
elif eType in ("Genotype", "Tissue", "Development_Phase"):
return ("Biological_context", "Context")
elif eType == "Environmental_Factor":
return ("Context",)
else:
raise Exception("Unknown SeeDev type '" + str(eType) + "'")
def mergeForSeeDev(self, categoryName, structureAnalyzer):
if categoryName in structureAnalyzer.typeMap["forward"]:
return structureAnalyzer.typeMap["forward"][categoryName]
return categoryName
# for tag in ("Regulates", "Exists", "Interacts", "Is", "Occurs"):
# if categoryName.startswith(tag):
# categoryName = tag
# break
# return categoryName
def processCorpus(self, input, output, gold=None, append=False, allowNewIds=True, structureAnalyzer=None):
if self.styles["sdb_merge"]:
structureAnalyzer.determineNonOverlappingTypes()
self.structureAnalyzer = structureAnalyzer
ExampleBuilder.processCorpus(self, input, output, gold, append, allowNewIds, structureAnalyzer)
def isValidInteraction(self, e1, e2, structureAnalyzer,forceUndirected=False):
return len(structureAnalyzer.getValidEdgeTypes(e1.get("type"), e2.get("type"), forceUndirected=forceUndirected)) > 0
def getGoldCategoryName(self, goldGraph, entityToGold, e1, e2, directed=True):
if len(entityToGold[e1]) > 0 and len(entityToGold[e2]) > 0:
return self.getCategoryName(goldGraph, entityToGold[e1][0], entityToGold[e2][0], directed=directed)
else:
return "neg"
def filterEdge(self, edge, edgeTypes):
import types
assert edgeTypes != None
if type(edgeTypes) not in [types.ListType, types.TupleType]:
edgeTypes = [edgeTypes]
if edge[2].get("type") in edgeTypes:
return True
else:
return False
def keepExample(self, e1, e2, categoryName, isDirected, structureAnalyzer):
makeExample = True
if (not self.styles["no_auto_limits"]) and not self.isValidInteraction(e1, e2, structureAnalyzer, forceUndirected=not isDirected):
makeExample = False
self.exampleStats.filter("auto_limits")
if self.styles["genia_task1"] and (e1.get("type") == "Entity" or e2.get("type") == "Entity"):
makeExample = False
self.exampleStats.filter("genia_task1")
if self.styles["pos_only"] and categoryName == "neg":
makeExample = False
self.exampleStats.filter("pos_only")
if self.styles["no_self_loops"] and ((e1 == e2) or (e1.get("headOffset") == e2.get("headOffset"))):
makeExample = False
self.exampleStats.filter("no_self_loops")
return makeExample
def getExampleCategoryName(self, e1=None, e2=None, t1=None, t2=None, sentenceGraph=None, goldGraph=None, entityToGold=None, isDirected=True, structureAnalyzer=None):
if self.styles["token_nodes"]:
categoryName = self.getCategoryNameFromTokens(sentenceGraph, t1, t2, isDirected)
else:
categoryName = self.getCategoryName(sentenceGraph, e1, e2, isDirected)
if goldGraph != None:
categoryName = self.getGoldCategoryName(goldGraph, entityToGold, e1, e2, isDirected)
if self.styles["filter_types"] != None and categoryName in self.styles["filter_types"]:
categoryName = "neg"
if self.styles["se10t8_undirected"]:
assert e1.get("id").endswith(".e1")
assert e2.get("id").endswith(".e2")
#if self.styles["sdb_merge"]:
# categoryName = self.mergeForSeeDev(categoryName, structureAnalyzer)
return categoryName
def buildExamplesFromGraph(self, sentenceGraph, outfile, goldGraph = None, structureAnalyzer=None):
"""
Build examples for a single sentence. Returns a list of examples.
See Core/ExampleUtils for example format.
"""
#examples = []
exampleIndex = 0
# example directionality
if self.styles["directed"] == None and self.styles["undirected"] == None: # determine directedness from corpus
examplesAreDirected = structureAnalyzer.hasDirectedTargets() if structureAnalyzer != None else True
elif self.styles["directed"]:
assert self.styles["undirected"] in [None, False]
examplesAreDirected = True
elif self.styles["undirected"]:
assert self.styles["directed"] in [None, False]
examplesAreDirected = False
if not self.styles["no_trigger_features"]:
self.triggerFeatureBuilder.initSentence(sentenceGraph)
if self.styles["evex"]:
self.evexFeatureBuilder.initSentence(sentenceGraph)
# if self.styles["sdb_merge"]:
# self.determineNonOverlappingTypes(structureAnalyzer)
# Filter entities, if needed
sentenceGraph.mergeInteractionGraph(True)
entities = sentenceGraph.mergedEntities
entityToDuplicates = sentenceGraph.mergedEntityToDuplicates
self.exampleStats.addValue("Duplicate entities skipped", len(sentenceGraph.entities) - len(entities))
# Connect to optional gold graph
entityToGold = None
if goldGraph != None:
entityToGold = EvaluateInteractionXML.mapEntities(entities, goldGraph.entities)
paths = None
if not self.styles["no_path"]:
undirected = sentenceGraph.dependencyGraph.toUndirected()
paths = undirected
if self.styles["filter_shortest_path"] != None: # For DDI use filter_shortest_path=conj_and
paths.resetAnalyses() # just in case
paths.FloydWarshall(self.filterEdge, {"edgeTypes":self.styles["filter_shortest_path"]})
# Generate examples based on interactions between entities or interactions between tokens
if self.styles["token_nodes"]:
loopRange = len(sentenceGraph.tokens)
else:
loopRange = len(entities)
for i in range(loopRange-1):
for j in range(i+1,loopRange):
eI = None
eJ = None
if self.styles["token_nodes"]:
tI = sentenceGraph.tokens[i]
tJ = sentenceGraph.tokens[j]
else:
eI = entities[i]
eJ = entities[j]
tI = sentenceGraph.entityHeadTokenByEntity[eI]
tJ = sentenceGraph.entityHeadTokenByEntity[eJ]
if eI.get("type") == "neg" or eJ.get("type") == "neg":
continue
if self.styles["skip_extra_triggers"]:
if eI.get("source") != None or eJ.get("source") != None:
continue
# only consider paths between entities (NOTE! entities, not only named entities)
if self.styles["headsOnly"]:
if (len(sentenceGraph.tokenIsEntityHead[tI]) == 0) or (len(sentenceGraph.tokenIsEntityHead[tJ]) == 0):
continue
examples = self.buildExamplesForPair(tI, tJ, paths, sentenceGraph, goldGraph, entityToGold, eI, eJ, structureAnalyzer, examplesAreDirected)
for categoryName, features, extra in examples:
# make example
if self.styles["binary"]:
if categoryName != "neg":
category = 1
else:
category = -1
extra["categoryName"] = "i"
else:
category = self.classSet.getId(categoryName)
example = [sentenceGraph.getSentenceId()+".x"+str(exampleIndex), category, features, extra]
ExampleUtils.appendExamples([example], outfile)
exampleIndex += 1
return exampleIndex
def buildExamplesForPair(self, token1, token2, paths, sentenceGraph, goldGraph, entityToGold, entity1=None, entity2=None, structureAnalyzer=None, isDirected=True):
# define forward
categoryName = self.getExampleCategoryName(entity1, entity2, token1, token2, sentenceGraph, goldGraph, entityToGold, isDirected, structureAnalyzer=structureAnalyzer)
# make forward
forwardExample = None
self.exampleStats.beginExample(categoryName)
if self.keepExample(entity1, entity2, categoryName, isDirected, structureAnalyzer):
forwardExample = self.buildExample(token1, token2, paths, sentenceGraph, categoryName, entity1, entity2, structureAnalyzer, isDirected)
if isDirected: # build a separate reverse example (if that is valid)
self.exampleStats.endExample() # end forward example
# define reverse
categoryName = self.getExampleCategoryName(entity2, entity1, token2, token1, sentenceGraph, goldGraph, entityToGold, True, structureAnalyzer=structureAnalyzer)
# make reverse
self.exampleStats.beginExample(categoryName)
reverseExample = None
if self.keepExample(entity2, entity1, categoryName, True, structureAnalyzer):
reverseExample = self.buildExample(token2, token1, paths, sentenceGraph, categoryName, entity2, entity1, structureAnalyzer, isDirected)
self.exampleStats.endExample()
return filter(None, [forwardExample, reverseExample])
elif self.styles["se10t8_undirected"]: # undirected example with a directed type
self.exampleStats.endExample()
return [forwardExample]
elif forwardExample != None: # merge features from the reverse example to the forward one
reverseExample = self.buildExample(token2, token1, paths, sentenceGraph, categoryName, entity2, entity1, structureAnalyzer, isDirected)
forwardExample[1].update(reverseExample[1])
self.exampleStats.endExample() # end merged example
return [forwardExample]
else: # undirected example that was filtered
self.exampleStats.endExample() # end merged example
return []
def buildExample(self, token1, token2, paths, sentenceGraph, categoryName, entity1=None, entity2=None, structureAnalyzer=None, isDirected=True):
"""
Build a single directed example for the potential edge between token1 and token2
"""
# define features
if not self.styles["no_path"]:
path = paths.getPaths(token1, token2)
if len(path) > 0:
path = path[0]
#pathExists = True
else:
path = [token1, token2]
#pathExists = False
else:
path = [token1, token2]
#pathExists = False
features = {}
if not self.styles["no_features"]:
features = self.buildFeatures(sentenceGraph, entity1, entity2, token1, token2, path)
# define extra attributes
if int(path[0].get("charOffset").split("-")[0]) < int(path[-1].get("charOffset").split("-")[0]):
extra = {"xtype":"edge","type":"i","t1":path[0].get("id"),"t2":path[-1].get("id")}
extra["deprev"] = False
else:
extra = {"xtype":"edge","type":"i","t1":path[-1].get("id"),"t2":path[0].get("id")}
extra["deprev"] = True
if entity1 != None:
extra["e1"] = entity1.get("id")
if sentenceGraph.mergedEntityToDuplicates != None:
extra["e1DuplicateIds"] = ",".join([x.get("id") for x in sentenceGraph.mergedEntityToDuplicates[entity1]])
if entity2 != None:
extra["e2"] = entity2.get("id")
if sentenceGraph.mergedEntityToDuplicates != None:
extra["e2DuplicateIds"] = ",".join([x.get("id") for x in sentenceGraph.mergedEntityToDuplicates[entity2]])
extra["categoryName"] = categoryName
if self.styles["bacteria_renaming"]:
if entity1.get("text") != None and entity1.get("text") != "":
extra["e1t"] = entity1.get("text").replace(" ", "---").replace(":","-COL-")
if entity2.get("text") != None and entity2.get("text") != "":
extra["e2t"] = entity2.get("text").replace(" ", "---").replace(":","-COL-")
if self.styles["doc_extra"]:
if hasattr(sentenceGraph, "documentElement") and sentenceGraph.documentElement.get("origId") != None:
extra["DOID"] = sentenceGraph.documentElement.get("origId")
if self.styles["entity_extra"]:
if entity1.get("origId") != None: extra["e1OID"] = entity1.get("origId")
if entity2.get("origId") != None: extra["e2OID"] = entity2.get("origId")
sentenceOrigId = sentenceGraph.sentenceElement.get("origId")
if sentenceOrigId != None:
extra["SOID"] = sentenceOrigId
extra["directed"] = str(isDirected)
if self.styles["sdb_merge"]:
extra["sdb_merge"] = "True"
#print extra
return (categoryName, features, extra)
def buildFeatures(self, sentenceGraph, entity1, entity2, token1, token2, path):
features = {}
if not self.styles["no_trigger_features"]: # F 85.52 -> 85.55
self.triggerFeatureBuilder.setFeatureVector(features)
self.triggerFeatureBuilder.tag = "trg1_"
self.triggerFeatureBuilder.buildFeatures(token1)
self.triggerFeatureBuilder.tag = "trg2_"
self.triggerFeatureBuilder.buildFeatures(token2)
self.triggerFeatureBuilder.setFeatureVector(None)
# REL features
if self.styles["rel_features"] and not self.styles["no_task"]:
self.relFeatureBuilder.setFeatureVector(features)
self.relFeatureBuilder.tag = "rel1_"
self.relFeatureBuilder.buildAllFeatures(sentenceGraph.tokens, sentenceGraph.tokens.index(token1))
self.relFeatureBuilder.tag = "rel2_"
self.relFeatureBuilder.buildAllFeatures(sentenceGraph.tokens, sentenceGraph.tokens.index(token2))
self.relFeatureBuilder.setFeatureVector(None)
if self.styles["bacteria_renaming"] and not self.styles["no_task"]:
self.bacteriaRenamingFeatureBuilder.setFeatureVector(features)
self.bacteriaRenamingFeatureBuilder.buildPairFeatures(entity1, entity2)
#self.bacteriaRenamingFeatureBuilder.buildSubstringFeatures(entity1, entity2) # decreases perf. 74.76 -> 72.41
self.bacteriaRenamingFeatureBuilder.setFeatureVector(None)
if self.styles["co_features"] and not self.styles["no_task"]:
e1Offset = Range.charOffsetToSingleTuple(entity1.get("charOffset"))
e2Offset = Range.charOffsetToSingleTuple(entity2.get("charOffset"))
if Range.contains(e1Offset, e2Offset):
features[self.featureSet.getId("e1_contains_e2")] = 1
if entity2.get("given") == "True":
features[self.featureSet.getId("e1_contains_e2name")] = 1
if Range.contains(e2Offset, e1Offset):
features[self.featureSet.getId("e2_contains_e1")] = 1
if entity1.get("given") == "True":
features[self.featureSet.getId("e2_contains_e1name")] = 1
if self.styles["drugbank_features"]:
self.drugFeatureBuilder.setFeatureVector(features)
self.drugFeatureBuilder.tag = "ddi_"
self.drugFeatureBuilder.buildPairFeatures(entity1, entity2)
if self.styles["ddi_mtmx"]:
self.drugFeatureBuilder.buildMTMXFeatures(entity1, entity2)
self.drugFeatureBuilder.setFeatureVector(None)
if self.styles["graph_kernel"]:
self.graphKernelFeatureBuilder.setFeatureVector(features, entity1, entity2)
self.graphKernelFeatureBuilder.buildGraphKernelFeatures(sentenceGraph, path)
self.graphKernelFeatureBuilder.setFeatureVector(None)
if self.styles["entity_type"]:
e1Type = self.multiEdgeFeatureBuilder.getEntityType(entity1)
e2Type = self.multiEdgeFeatureBuilder.getEntityType(entity2)
features[self.featureSet.getId("e1_"+e1Type)] = 1
features[self.featureSet.getId("e2_"+e2Type)] = 1
features[self.featureSet.getId("distance_"+str(len(path)))] = 1
if not self.styles["no_dependency"]:
#print "Dep features"
self.multiEdgeFeatureBuilder.setFeatureVector(features, entity1, entity2)
#self.multiEdgeFeatureBuilder.buildStructureFeatures(sentenceGraph, paths) # remove for fast
if not self.styles["disable_entity_features"]:
self.multiEdgeFeatureBuilder.buildEntityFeatures(sentenceGraph)
self.multiEdgeFeatureBuilder.buildPathLengthFeatures(path)
if not self.styles["disable_terminus_features"]:
self.multiEdgeFeatureBuilder.buildTerminusTokenFeatures(path, sentenceGraph) # remove for fast
if not self.styles["disable_single_element_features"]:
self.multiEdgeFeatureBuilder.buildSingleElementFeatures(path, sentenceGraph)
if not self.styles["disable_ngram_features"]:
#print "NGrams"
self.multiEdgeFeatureBuilder.buildPathGrams(2, path, sentenceGraph) # remove for fast
self.multiEdgeFeatureBuilder.buildPathGrams(3, path, sentenceGraph) # remove for fast
self.multiEdgeFeatureBuilder.buildPathGrams(4, path, sentenceGraph) # remove for fast
#self.buildEdgeCombinations(path, edges, sentenceGraph, features) # remove for fast
#if edges != None:
# self.multiEdgeFeatureBuilder.buildTerminusFeatures(path[0], edges[0][1]+edges[1][0], "t1", sentenceGraph) # remove for fast
# self.multiEdgeFeatureBuilder.buildTerminusFeatures(path[-1], edges[len(path)-1][len(path)-2]+edges[len(path)-2][len(path)-1], "t2", sentenceGraph) # remove for fast
if not self.styles["disable_path_edge_features"]:
self.multiEdgeFeatureBuilder.buildPathEdgeFeatures(path, sentenceGraph)
self.multiEdgeFeatureBuilder.buildSentenceFeatures(sentenceGraph)
self.multiEdgeFeatureBuilder.setFeatureVector(None)
if self.styles["nodalida"]:
self.nodalidaFeatureBuilder.setFeatureVector(features, entity1, entity2)
shortestPaths = self.nodalidaFeatureBuilder.buildShortestPaths(sentenceGraph.dependencyGraph, path)
print shortestPaths
if len(shortestPaths) > 0:
self.nodalidaFeatureBuilder.buildNGrams(shortestPaths, sentenceGraph)
self.nodalidaFeatureBuilder.setFeatureVector(None)
if self.styles["linear_features"]:
self.tokenFeatureBuilder.setFeatureVector(features)
for i in range(len(sentenceGraph.tokens)):
if sentenceGraph.tokens[i] == token1:
token1Index = i
if sentenceGraph.tokens[i] == token2:
token2Index = i
linearPreTag = "linfw_"
if token1Index > token2Index:
token1Index, token2Index = token2Index, token1Index
linearPreTag = "linrv_"
self.tokenFeatureBuilder.buildLinearOrderFeatures(token1Index, sentenceGraph, 2, 2, preTag="linTok1")
self.tokenFeatureBuilder.buildLinearOrderFeatures(token2Index, sentenceGraph, 2, 2, preTag="linTok2")
# Before, middle, after
# self.tokenFeatureBuilder.buildTokenGrams(0, token1Index-1, sentenceGraph, "bf")
# self.tokenFeatureBuilder.buildTokenGrams(token1Index+1, token2Index-1, sentenceGraph, "bw")
# self.tokenFeatureBuilder.buildTokenGrams(token2Index+1, len(sentenceGraph.tokens)-1, sentenceGraph, "af")
# before-middle, middle, middle-after
# self.tokenFeatureBuilder.buildTokenGrams(0, token2Index-1, sentenceGraph, linearPreTag+"bf", max=2)
# self.tokenFeatureBuilder.buildTokenGrams(token1Index+1, token2Index-1, sentenceGraph, linearPreTag+"bw", max=2)
# self.tokenFeatureBuilder.buildTokenGrams(token1Index+1, len(sentenceGraph.tokens)-1, sentenceGraph, linearPreTag+"af", max=2)
self.tokenFeatureBuilder.setFeatureVector(None)
if self.styles["random"]:
self.randomFeatureBuilder.setFeatureVector(features)
self.randomFeatureBuilder.buildRandomFeatures(100, 0.01)
self.randomFeatureBuilder.setFeatureVector(None)
if self.styles["genia_features"] and not self.styles["no_task"]:
e1Type = entity1.get("type")
e2Type = entity2.get("type")
assert(entity1.get("given") in (None, "False"))
if entity2.get("given") == "True":
features[self.featureSet.getId("GENIA_target_protein")] = 1
else:
features[self.featureSet.getId("GENIA_nested_event")] = 1
if e1Type.find("egulation") != -1: # leave r out to avoid problems with capitalization
if entity2.get("given") == "True":
features[self.featureSet.getId("GENIA_regulation_of_protein")] = 1
else:
features[self.featureSet.getId("GENIA_regulation_of_event")] = 1
if self.styles["bi_features"]:
# Make features based on entity types
e1Type = entity1.get("type")
e2Type = entity2.get("type")
e1SuperType = str(self.getBISuperType(e1Type))
e2SuperType = str(self.getBISuperType(e2Type))
features[self.featureSet.getId("BI_e1_"+e1Type)] = 1
features[self.featureSet.getId("BI_e2_"+e2Type)] = 1
features[self.featureSet.getId("BI_e1sup_"+e1SuperType)] = 1
features[self.featureSet.getId("BI_e2sup_"+e2SuperType)] = 1
features[self.featureSet.getId("BI_e1e2_"+e1Type+"_"+e2Type)] = 1
features[self.featureSet.getId("BI_e1e2sup_"+e1SuperType+"_"+e2SuperType)] = 1
if self.styles["sdb_features"]:
e1Type = entity1.get("type")
e2Type = entity2.get("type")
features[self.featureSet.getId("SDB_e1_"+e1Type)] = 1
features[self.featureSet.getId("SDB_e2_"+e2Type)] = 1
features[self.featureSet.getId("SDB_e1e2_"+e1Type+"_"+e2Type)] = 1
if e1Type == e2Type:
features[self.featureSet.getId("SDB_e1e2_equal")] = 1
features[self.featureSet.getId("SDB_e1e2_equal_" + e1Type)] = 1
e1SuperTypes = str(self.getSeeDevSuperTypes(e1Type))
e2SuperTypes = str(self.getSeeDevSuperTypes(e2Type))
for e1SuperType in e1SuperTypes:
for e2SuperType in e2SuperTypes:
features[self.featureSet.getId("SDB_e1sup_"+e1SuperType)] = 1
features[self.featureSet.getId("SDB_e2sup_"+e2SuperType)] = 1
features[self.featureSet.getId("SDB_e1e2sup_"+e1SuperType+"_"+e2SuperType)] = 1
if e1SuperType == e2SuperType:
features[self.featureSet.getId("SDB_e1e2sup_equal")] = 1
features[self.featureSet.getId("SDB_e1e2sup_equal_" + e1SuperType)] = 1
if self.styles["ontobiotope_features"]:
self.ontobiotopeFeatureBuilder.setFeatureVector(features)
self.ontobiotopeFeatureBuilder.buildOBOFeaturesForEntityPair(entity1, entity2)
self.ontobiotopeFeatureBuilder.setFeatureVector(None)
if self.styles["full_entities"]:
e1Text = entity1.get("text").lower()
e2Text = entity2.get("text").lower()
features[self.featureSet.getId("FULL_e1_"+e1Text)] = 1
features[self.featureSet.getId("FULL_e2_"+e2Text)] = 1
for ep1 in e1Text.split():
for ep2 in e2Text.split():
features[self.featureSet.getId("FULL_e1_"+ep1)] = 1
features[self.featureSet.getId("FULL_e2_"+ep2)] = 1
features[self.featureSet.getId("FULL_e1e2_"+ep1+"_"+ep2)] = 1
if self.styles["evex"]:
self.evexFeatureBuilder.setFeatureVector(features, entity1, entity2)
self.evexFeatureBuilder.buildEdgeFeatures(entity1, entity2, token1, token2, path, sentenceGraph)
self.evexFeatureBuilder.setFeatureVector(None)
if self.styles["wordnet"]:
self.wordNetFeatureBuilder.setFeatureVector(features, entity1, entity2)
self.wordNetFeatureBuilder.buildFeaturesForEntityPair(token1, token2)
self.wordNetFeatureBuilder.buildLinearFeatures(token1, sentenceGraph.tokens, tag="t1_")
self.wordNetFeatureBuilder.buildLinearFeatures(token2, sentenceGraph.tokens, tag="t2_")
self.wordNetFeatureBuilder.buildPathFeatures(path)
self.wordNetFeatureBuilder.setFeatureVector(None)
if self.styles["wordvector"]:
self.wordVectorFeatureBuilder.setFeatureVector(features, entity1, entity2)
self.wordVectorFeatureBuilder.buildFeatures(token1, "t1_")
self.wordVectorFeatureBuilder.buildFeatures(token2, "t2_")
self.wordVectorFeatureBuilder.buildLinearFeatures(token1, sentenceGraph.tokens, tag="t1_")
self.wordVectorFeatureBuilder.buildLinearFeatures(token2, sentenceGraph.tokens, tag="t2_")
self.wordVectorFeatureBuilder.buildPathFeatures(path)
self.wordVectorFeatureBuilder.buildFBAFeatures(sentenceGraph.tokens, sentenceGraph.tokens.index(token1), sentenceGraph.tokens.index(token2))
self.wordVectorFeatureBuilder.setFeatureVector(None)
if self.styles["giuliano"]:
self.giulianoFeatureBuilder.setFeatureVector(features, entity1, entity2)
self.giulianoFeatureBuilder.buildEdgeFeatures(entity1, entity2, token1, token2, path, sentenceGraph)
self.giulianoFeatureBuilder.setFeatureVector(None)
return features
class GeneralEntityTypeRecognizerGztr(ExampleBuilder):
def __init__(self,
style=None,
classSet=None,
featureSet=None,
gazetteerFileName=None,
skiplist=None):
if classSet == None:
classSet = IdSet(1)
assert (classSet.getId("neg") == 1)
if featureSet == None:
featureSet = IdSet()
ExampleBuilder.__init__(self, classSet, featureSet)
#gazetteerFileName="/usr/share/biotext/GeniaChallenge/SharedTaskTriggerTest/gazetteer-train"
if gazetteerFileName != None:
self.gazetteer = Gazetteer.loadGztr(gazetteerFileName)
print >> sys.stderr, "Loaded gazetteer from", gazetteerFileName
else:
print >> sys.stderr, "No gazetteer loaded"
self.gazetteer = None
self.styles = self.getParameters(style, [
"rel_features", "wordnet", "bb_features", "giuliano",
"epi_merge_negated", "limit_merged_types", "genia_task1",
"build_for_nameless", "pos_only", "all_tokens", "names",
"pos_pairs", "linear_ngrams", "phospho"
])
# if "selftrain_group" in self.styles:
# self.selfTrainGroups = set()
# if "selftrain_group-1" in self.styles:
# self.selfTrainGroups.add("-1")
# if "selftrain_group0" in self.styles:
# self.selfTrainGroups.add("0")
# if "selftrain_group1" in self.styles:
# self.selfTrainGroups.add("1")
# if "selftrain_group2" in self.styles:
# self.selfTrainGroups.add("2")
# if "selftrain_group3" in self.styles:
# self.selfTrainGroups.add("3")
# print >> sys.stderr, "Self-train-groups:", self.selfTrainGroups
self.skiplist = set()
if skiplist != None:
f = open(skiplist, "rt")
for line in f.readlines():
self.skiplist.add(line.strip())
f.close()
if self.styles["rel_features"]:
self.relFeatureBuilder = RELFeatureBuilder(featureSet)
if self.styles["wordnet"]:
self.wordNetFeatureBuilder = WordNetFeatureBuilder(featureSet)
if self.styles["bb_features"]:
self.bacteriaTokens = PhraseTriggerExampleBuilder.getBacteriaTokens(
PhraseTriggerExampleBuilder.getBacteriaNames())
if self.styles["giuliano"]:
self.giulianoFeatureBuilder = GiulianoFeatureBuilder(featureSet)
def getMergedEntityType(self, entities):
"""
If a single token belongs to multiple entities of different types,
a new, composite type is defined. This type is the alphabetically
ordered types of these entities joined with '---'.
"""
types = set()
entityIds = set()
for entity in entities:
if entity.get("isName") == "True" and self.styles["all_tokens"]:
continue
if entity.get("type") == "Entity" and self.styles["genia_task1"]:
continue
if self.styles["epi_merge_negated"]:
types.add(
InteractionXML.ResolveEPITriggerTypes.getEPIBaseType(
entity.get("type")))
entityIds.add(entity.get("id"))
else:
types.add(entity.get("type"))
entityIds.add(entity.get("id"))
types = list(types)
types.sort()
typeString = ""
for type in types:
#if type == "Protein" and "all_tokens" in self.styles:
# continue
if typeString != "":
typeString += "---"
typeString += type
if typeString == "":
return "neg", None
idString = "/".join(sorted(list(entityIds)))
if self.styles["limit_merged_types"]:
if typeString.find("---") != -1:
if typeString == "Gene_expression---Positive_regulation":
return typeString, idString
else:
return typeString.split(
"---")[0], idString # ids partially incorrect
else:
return typeString, idString
return typeString, idString
def getTokenFeatures(self, token, sentenceGraph):
"""
Returns a list of features based on the attributes of a token.
These can be used to define more complex features.
"""
# These features are cached when this method is first called
# for a token.
if self.tokenFeatures.has_key(token):
return self.tokenFeatures[token], self.tokenFeatureWeights[token]
tokTxt = sentenceGraph.getTokenText(token)
features = {}
features["_txt_" + tokTxt] = 1
features["_POS_" + token.get("POS")] = 1
if sentenceGraph.tokenIsName[token]:
features["_isName"] = 1
for entity in sentenceGraph.tokenIsEntityHead[token]:
if entity.get("isName") == "True":
features["_annType_" + entity.get("type")] = 1
# # Filip's gazetteer based features (can be used separately from exclude_gazetteer)
# if "gazetteer_features" in self.styles:
# tokTxtLower = tokTxt.lower()
# if "stem_gazetteer" in self.styles:
# tokTxtLower = PorterStemmer.stem(tokTxtLower)
# if self.gazetteer and tokTxtLower in self.gazetteer:
# for label,weight in self.gazetteer[tokTxtLower].items():
# features["_knownLabel_"+label]=weight # 1 performs slightly worse
## BANNER features
#if sentenceGraph.entityHintsByToken.has_key(token):
# features["BANNER-entity"] = 1
# Wordnet features
#if "wordnet" in self.styles:
# for wordNetFeature in self.wordNetFeatureBuilder.getTokenFeatures(tokTxt, token.get("POS")):
# features["_WN_"+wordNetFeature] = 1
self.tokenFeatures[token] = sorted(features.keys())
self.tokenFeatureWeights[token] = features
return self.tokenFeatures[token], self.tokenFeatureWeights[token]
def buildLinearOrderFeatures(self, sentenceGraph, index, tag, features):
"""
Linear features are built by marking token features with a tag
that defines their relative position in the linear order.
"""
tag = "linear_" + tag
tokenFeatures, tokenFeatureWeights = self.getTokenFeatures(
sentenceGraph.tokens[index], sentenceGraph)
for tokenFeature in tokenFeatures:
features[self.featureSet.getId(
tag + tokenFeature)] = tokenFeatureWeights[tokenFeature]
def buildLinearNGram(self, i, j, sentenceGraph, features):
ngram = "ngram"
for index in range(i, j + 1):
ngram += "_" + sentenceGraph.getTokenText(
sentenceGraph.tokens[index]).lower()
features[self.featureSet.getId(ngram)] = 1
def buildExamplesFromGraph(self, sentenceGraph, outfile, goldGraph=None):
"""
Build one example for each token of the sentence
"""
if sentenceGraph.sentenceElement.get("origId") in self.skiplist:
print >> sys.stderr, "Skipping sentence", sentenceGraph.sentenceElement.get(
"origId")
return 0 #[]
#examples = []
exampleIndex = 0
self.tokenFeatures = {}
self.tokenFeatureWeights = {}
namedEntityHeadTokens = []
if not self.styles["names"]:
namedEntityCount = 0
for entity in sentenceGraph.entities:
if entity.get(
"isName"
) == "True": # known data which can be used for features
namedEntityCount += 1
namedEntityCountFeature = "nameCount_" + str(namedEntityCount)
# NOTE!!! This will change the number of examples and omit
# all triggers (positive and negative) from sentences which
# have no NE:s, possibly giving a too-optimistic performance
# value. Such sentences can still have triggers from intersentence
# interactions, but as such events cannot be recovered anyway,
# looking for these triggers would be pointless.
if namedEntityCount == 0 and not self.styles[
"build_for_nameless"]: # no names, no need for triggers
return 0 #[]
if self.styles["pos_pairs"]:
namedEntityHeadTokens = self.getNamedEntityHeadTokens(
sentenceGraph)
bagOfWords = {}
for token in sentenceGraph.tokens:
text = "bow_" + token.get("text")
if not bagOfWords.has_key(text):
bagOfWords[text] = 0
bagOfWords[text] += 1
if sentenceGraph.tokenIsName[token]:
text = "ne_" + text
if not bagOfWords.has_key(text):
bagOfWords[text] = 0
bagOfWords[text] += 1
bowFeatures = {}
for k in sorted(bagOfWords.keys()):
bowFeatures[self.featureSet.getId(k)] = bagOfWords[k]
self.inEdgesByToken = {}
self.outEdgesByToken = {}
self.edgeSetByToken = {}
for token in sentenceGraph.tokens:
#inEdges = sentenceGraph.dependencyGraph.in_edges(token, data=True)
#fixedInEdges = []
#for edge in inEdges:
# fixedInEdges.append( (edge[0], edge[1], edge[2]["element"]) )
#inEdges = fixedInEdges
inEdges = sentenceGraph.dependencyGraph.getInEdges(token)
#inEdges.sort(compareDependencyEdgesById)
self.inEdgesByToken[token] = inEdges
#outEdges = sentenceGraph.dependencyGraph.out_edges(token, data=True)
#fixedOutEdges = []
#for edge in outEdges:
# fixedOutEdges.append( (edge[0], edge[1], edge[2]["element"]) )
#outEdges = fixedOutEdges
outEdges = sentenceGraph.dependencyGraph.getOutEdges(token)
#outEdges.sort(compareDependencyEdgesById)
self.outEdgesByToken[token] = outEdges
self.edgeSetByToken[token] = set(inEdges + outEdges)
for i in range(len(sentenceGraph.tokens)):
token = sentenceGraph.tokens[i]
# CLASS
if len(sentenceGraph.tokenIsEntityHead[token]) > 0:
categoryName, entityIds = self.getMergedEntityType(
sentenceGraph.tokenIsEntityHead[token])
else:
categoryName, entityIds = "neg", None
self.exampleStats.beginExample(categoryName)
# Recognize only non-named entities (i.e. interaction words)
if sentenceGraph.tokenIsName[token] and not self.styles[
"names"] and not self.styles["all_tokens"]:
self.exampleStats.filter("name")
self.exampleStats.endExample()
continue
# if "selftrain_limits" in self.styles:
# # any predicted entity not part of the self-training set causes example to be rejected
# filtered = False
# for entity in sentenceGraph.tokenIsEntityHead[token]:
# if entity.get("selftrain") == "False":
# self.exampleStats.filter("selftrain_limits")
# self.exampleStats.endExample()
# filtered = True
# break
# if filtered:
# continue
# if "selftrain_group" in self.styles:
# # any predicted entity not part of the self-training set causes example to be rejected
# filtered = False
# for entity in sentenceGraph.tokenIsEntityHead[token]:
# if entity.get("selftraingroup") not in self.selfTrainGroups:
# self.exampleStats.filter("selftrain_group")
# self.exampleStats.endExample()
# filtered = True
# break
# if filtered:
# continue
if self.styles["pos_only"] and categoryName == "neg":
self.exampleStats.filter("pos_only")
self.exampleStats.endExample()
continue
category = self.classSet.getId(categoryName)
tokenText = token.get("text").lower()
# if "stem_gazetteer" in self.styles:
# tokenText = PorterStemmer.stem(tokenText)
# if ("exclude_gazetteer" in self.styles) and self.gazetteer and tokenText not in self.gazetteer:
# features = {}
# features[self.featureSet.getId("exclude_gazetteer")] = 1
# extra = {"xtype":"token","t":token.get("id"),"excluded":"True"}
# if entityIds != None:
# extra["goldIds"] = entityIds
# #examples.append( (sentenceGraph.getSentenceId()+".x"+str(exampleIndex),category,features,extra) )
# ExampleUtils.appendExamples([(sentenceGraph.getSentenceId()+".x"+str(exampleIndex),category,features,extra)], outfile)
# exampleIndex += 1
# continue
# FEATURES
features = {}
if not self.styles["names"]:
features[self.featureSet.getId(namedEntityCountFeature)] = 1
#for k,v in bagOfWords.iteritems():
# features[self.featureSet.getId(k)] = v
# pre-calculate bow _features_
features.update(bowFeatures)
# for j in range(len(sentenceGraph.tokens)):
# text = "bow_" + sentenceGraph.tokens[j].get("text")
# if j < i:
# features[self.featureSet.getId("bf_" + text)] = 1
# elif j > i:
# features[self.featureSet.getId("af_" + text)] = 1
# Main features
text = token.get("text")
features[self.featureSet.getId("txt_" + text)] = 1
features[self.featureSet.getId("POS_" + token.get("POS"))] = 1
stem = PorterStemmer.stem(text)
features[self.featureSet.getId("stem_" + stem)] = 1
features[self.featureSet.getId("nonstem_" + text[len(stem):])] = 1
# Normalized versions of the string (if same as non-normalized, overlap without effect)
normalizedText = text.replace("-", "").replace("/", "").replace(
",", "").replace("\\", "").replace(" ", "").lower()
if normalizedText == "bound": # should be for all irregular verbs
normalizedText = "bind"
features[self.featureSet.getId("txt_" + normalizedText)] = 1
norStem = PorterStemmer.stem(normalizedText)
features[self.featureSet.getId("stem_" + norStem)] = 1
features[self.featureSet.getId("nonstem_" +
normalizedText[len(norStem):])] = 1
## Subspan features
#textLower = text.lower()
#for i in range(1, len(textLower)):
# features[self.featureSet.getId("subspanbegin"+str(i)+"_"+textLower[0:i])] = 1
# features[self.featureSet.getId("subspanend"+str(i)+"_"+textLower[-i:])] = 1
# Substring features
for string in text.split("-"):
stringLower = string.lower()
features[self.featureSet.getId("substring_" + stringLower)] = 1
features[self.featureSet.getId(
"substringstem_" + PorterStemmer.stem(stringLower))] = 1
# Linear order features
for index in [-3, -2, -1, 1, 2, 3]:
if i + index > 0 and i + index < len(sentenceGraph.tokens):
self.buildLinearOrderFeatures(sentenceGraph, i + index,
str(index), features)
# Linear n-grams
if self.styles["linear_ngrams"]:
self.buildLinearNGram(max(0, i - 1), i, sentenceGraph,
features)
self.buildLinearNGram(max(0, i - 2), i, sentenceGraph,
features)
if self.styles["phospho"]:
if text.find("hospho") != -1:
features[self.featureSet.getId("phospho_found")] = 1
features[self.featureSet.getId("begin_" +
text[0:2].lower())] = 1
features[self.featureSet.getId("begin_" +
text[0:3].lower())] = 1
if self.styles["bb_features"]:
if text.lower() in self.bacteriaTokens:
features[self.featureSet.getId("lpsnBacToken")] = 1
# Content
if i > 0 and text[0].isalpha() and text[0].isupper():
features[self.featureSet.getId("upper_case_start")] = 1
for j in range(len(text)):
if j > 0 and text[j].isalpha() and text[j].isupper():
features[self.featureSet.getId("upper_case_middle")] = 1
# numbers and special characters
if text[j].isdigit():
features[self.featureSet.getId("has_digits")] = 1
if j > 0 and text[j - 1] == "-":
features[self.featureSet.getId(
"has_hyphenated_digit")] = 1
elif text[j] == "-":
features[self.featureSet.getId("has_hyphen")] = 1
elif text[j] == "/":
features[self.featureSet.getId("has_fslash")] = 1
elif text[j] == "\\":
features[self.featureSet.getId("has_bslash")] = 1
# duplets
if j > 0:
features[self.featureSet.getId("dt_" +
text[j - 1:j +
1].lower())] = 1
# triplets
if j > 1:
features[self.featureSet.getId("tt_" +
text[j - 2:j +
1].lower())] = 1
# quadruplets (don't work, slight decrease (0.5 pp) on f-score
#if j > 2:
# features[self.featureSet.getId("qt_"+text[j-3:j+1].lower())] = 1
# Attached edges (Hanging in and out edges)
t1InEdges = self.inEdgesByToken[token]
for edge in t1InEdges:
edgeType = edge[2].get("type")
features[self.featureSet.getId("t1HIn_" + edgeType)] = 1
features[self.featureSet.getId("t1HIn_" +
edge[0].get("POS"))] = 1
features[self.featureSet.getId("t1HIn_" + edgeType + "_" +
edge[0].get("POS"))] = 1
tokenText = sentenceGraph.getTokenText(edge[0])
features[self.featureSet.getId("t1HIn_" + tokenText)] = 1
features[self.featureSet.getId("t1HIn_" + edgeType + "_" +
tokenText)] = 1
tokenStem = PorterStemmer.stem(tokenText)
features[self.featureSet.getId("t1HIn_" + tokenStem)] = 1
features[self.featureSet.getId("t1HIn_" + edgeType + "_" +
tokenStem)] = 1
features[self.featureSet.getId("t1HIn_" + norStem + "_" +
edgeType + "_" + tokenStem)] = 1
t1OutEdges = self.outEdgesByToken[token]
for edge in t1OutEdges:
edgeType = edge[2].get("type")
features[self.featureSet.getId("t1HOut_" + edgeType)] = 1
features[self.featureSet.getId("t1HOut_" +
edge[1].get("POS"))] = 1
features[self.featureSet.getId("t1HOut_" + edgeType + "_" +
edge[1].get("POS"))] = 1
tokenText = sentenceGraph.getTokenText(edge[1])
features[self.featureSet.getId("t1HOut_" + tokenText)] = 1
features[self.featureSet.getId("t1HOut_" + edgeType + "_" +
tokenText)] = 1
tokenStem = PorterStemmer.stem(tokenText)
features[self.featureSet.getId("t1HOut_" + tokenStem)] = 1
features[self.featureSet.getId("t1HOut_" + edgeType + "_" +
tokenStem)] = 1
features[self.featureSet.getId("t1HOut_" + norStem + "_" +
edgeType + "_" + tokenStem)] = 1
# REL features
if self.styles["rel_features"]:
self.relFeatureBuilder.setFeatureVector(features)
self.relFeatureBuilder.buildAllFeatures(
sentenceGraph.tokens, i)
self.relFeatureBuilder.setFeatureVector(None)
#self.wordNetFeatureBuilder.getTokenFeatures("show", "VBP")
#tokTxt = token.get("text")
#tokPOS = token.get("POS")
#wordNetFeatures = []
#wordNetFeatures = self.wordNetFeatureBuilder.getTokenFeatures(tokTxt, tokPOS)
#self.wordNetFeatureBuilder.getTokenFeatures(tokTxt, tokPOS)
if self.styles["wordnet"]:
tokTxt = token.get("text")
tokPOS = token.get("POS")
wordNetFeatures = self.wordNetFeatureBuilder.getTokenFeatures(
tokTxt, tokPOS)
for wordNetFeature in wordNetFeatures:
#print wordNetFeature,
features[self.featureSet.getId("WN_" + wordNetFeature)] = 1
#print
if self.styles["giuliano"]:
self.giulianoFeatureBuilder.setFeatureVector(features)
self.giulianoFeatureBuilder.buildTriggerFeatures(
token, sentenceGraph)
self.giulianoFeatureBuilder.setFeatureVector(None)
extra = {"xtype": "token", "t": token.get("id")}
if self.styles["bb_features"]:
extra[
"trigex"] = "bb" # Request trigger extension in ExampleWriter
if self.styles["epi_merge_negated"]:
extra["unmergeneg"] = "epi" # Request trigger type unmerging
if entityIds != None:
extra[
"goldIds"] = entityIds # The entities to which this example corresponds
#examples.append( (sentenceGraph.getSentenceId()+".x"+str(exampleIndex),category,features,extra) )
# chains
self.buildChains(token, sentenceGraph, features)
if self.styles["pos_pairs"]:
self.buildPOSPairs(token, namedEntityHeadTokens, features)
example = (sentenceGraph.getSentenceId() + ".x" +
str(exampleIndex), category, features, extra)
ExampleUtils.appendExamples([example], outfile)
exampleIndex += 1
self.exampleStats.endExample()
#return examples
return exampleIndex
def buildChains(self,
token,
sentenceGraph,
features,
depthLeft=3,
chain="",
visited=None):
if depthLeft == 0:
return
strDepthLeft = "dist_" + str(depthLeft)
if visited == None:
visited = set()
inEdges = self.inEdgesByToken[token]
outEdges = self.outEdgesByToken[token]
edgeSet = visited.union(self.edgeSetByToken[token])
for edge in inEdges:
if not edge in visited:
edgeType = edge[2].get("type")
features[self.featureSet.getId("dep_" + strDepthLeft +
edgeType)] = 1
nextToken = edge[0]
tokenFeatures, tokenWeights = self.getTokenFeatures(
nextToken, sentenceGraph)
for tokenFeature in tokenFeatures:
features[self.featureSet.getId(
strDepthLeft +
tokenFeature)] = tokenWeights[tokenFeature]
# for entity in sentenceGraph.tokenIsEntityHead[nextToken]:
# if entity.get("isName") == "True":
# features[self.featureSet.getId("name_dist_"+strDepthLeft)] = 1
# features[self.featureSet.getId("name_dist_"+strDepthLeft+entity.get("type"))] = 1
# features[self.featureSet.getId("POS_dist_"+strDepthLeft+nextToken.get("POS"))] = 1
# tokenText = sentenceGraph.getTokenText(nextToken)
# features[self.featureSet.getId("text_dist_"+strDepthLeft+tokenText)] = 1
if sentenceGraph.tokenIsName[nextToken]:
features[self.featureSet.getId("name_chain_dist_" +
strDepthLeft + chain +
"-frw_" + edgeType)] = 1
features[self.featureSet.getId("chain_dist_" + strDepthLeft +
chain + "-frw_" + edgeType)] = 1
self.buildChains(nextToken, sentenceGraph, features,
depthLeft - 1, chain + "-frw_" + edgeType,
edgeSet)
for edge in outEdges:
if not edge in visited:
edgeType = edge[2].get("type")
features[self.featureSet.getId("dep_dist_" + strDepthLeft +
edgeType)] = 1
nextToken = edge[1]
tokenFeatures, tokenWeights = self.getTokenFeatures(
nextToken, sentenceGraph)
for tokenFeature in tokenFeatures:
features[self.featureSet.getId(
strDepthLeft +
tokenFeature)] = tokenWeights[tokenFeature]
# for entity in sentenceGraph.tokenIsEntityHead[nextToken]:
# if entity.get("isName") == "True":
# features[self.featureSet.getId("name_dist_"+strDepthLeft)] = 1
# features[self.featureSet.getId("name_dist_"+strDepthLeft+entity.get("type"))] = 1
# features[self.featureSet.getId("POS_dist_"+strDepthLeft+nextToken.get("POS"))] = 1
# tokenText = sentenceGraph.getTokenText(nextToken)
# features[self.featureSet.getId("text_dist_"+strDepthLeft+tokenText)] = 1
if sentenceGraph.tokenIsName[nextToken]:
features[self.featureSet.getId("name_chain_dist_" +
strDepthLeft + chain +
"-rev_" + edgeType)] = 1
features[self.featureSet.getId("chain_dist_" + strDepthLeft +
chain + "-rev_" + edgeType)] = 1
self.buildChains(nextToken, sentenceGraph, features,
depthLeft - 1, chain + "-rev_" + edgeType,
edgeSet)
def getNamedEntityHeadTokens(self, sentenceGraph):
headTokens = []
for entity in sentenceGraph.entities:
if entity.get(
"isName"
) == "True": # known data which can be used for features
headTokens.append(
sentenceGraph.entityHeadTokenByEntity[entity])
return headTokens
def buildPOSPairs(self, token, namedEntityHeadTokens, features):
tokenPOS = token.get("POS")
assert tokenPOS != None
for headToken in namedEntityHeadTokens:
headPOS = headToken.get("POS")
features[self.featureSet.getId("POS_pair_NE_" + tokenPOS + "-" +
headPOS)] = 1
def __init__(self,
style=None,
classSet=None,
featureSet=None,
gazetteerFileName=None,
skiplist=None):
if classSet == None:
classSet = IdSet(1)
if featureSet == None:
featureSet = IdSet()
ExampleBuilder.__init__(self, classSet, featureSet)
assert (classSet.getId("neg") == 1)
#gazetteerFileName="/usr/share/biotext/GeniaChallenge/SharedTaskTriggerTest/gazetteer-train"
if gazetteerFileName != None:
self.gazetteer = Gazetteer.loadGztr(gazetteerFileName)
print >> sys.stderr, "Loaded gazetteer from", gazetteerFileName
else:
print >> sys.stderr, "No gazetteer loaded"
self.gazetteer = None
self._setDefaultParameters([
"rel_features", "wordnet", "bb_features", "giuliano",
"epi_merge_negated", "limit_merged_types", "genia_task1", "names",
"build_for_nameless", "skip_for_nameless", "pos_only",
"all_tokens", "pos_pairs", "linear_ngrams", "phospho",
"drugbank_features", "ddi13_features", "metamap"
])
self.styles = self.getParameters(style)
# if "selftrain_group" in self.styles:
# self.selfTrainGroups = set()
# if "selftrain_group-1" in self.styles:
# self.selfTrainGroups.add("-1")
# if "selftrain_group0" in self.styles:
# self.selfTrainGroups.add("0")
# if "selftrain_group1" in self.styles:
# self.selfTrainGroups.add("1")
# if "selftrain_group2" in self.styles:
# self.selfTrainGroups.add("2")
# if "selftrain_group3" in self.styles:
# self.selfTrainGroups.add("3")
# print >> sys.stderr, "Self-train-groups:", self.selfTrainGroups
self.skiplist = set()
if skiplist != None:
f = open(skiplist, "rt")
for line in f.readlines():
self.skiplist.add(line.strip())
f.close()
if self.styles["rel_features"]:
self.relFeatureBuilder = RELFeatureBuilder(featureSet)
if self.styles["wordnet"]:
self.wordNetFeatureBuilder = WordNetFeatureBuilder(featureSet)
if self.styles["bb_features"]:
self.bacteriaTokens = PhraseTriggerExampleBuilder.getBacteriaTokens(
)
#self.bacteriaTokens = PhraseTriggerExampleBuilder.getBacteriaTokens(PhraseTriggerExampleBuilder.getBacteriaNames())
if self.styles["giuliano"]:
self.giulianoFeatureBuilder = GiulianoFeatureBuilder(featureSet)
if self.styles["drugbank_features"]:
self.drugFeatureBuilder = DrugFeatureBuilder(featureSet)
def __init__(self,
style=None,
length=None,
types=[],
featureSet=None,
classSet=None):
if featureSet == None:
featureSet = IdSet()
if classSet == None:
classSet = IdSet(1)
else:
classSet = classSet
assert (classSet.getId("neg") == 1
or (len(classSet.Ids) == 2 and classSet.getId("neg") == -1))
ExampleBuilder.__init__(self, classSet=classSet, featureSet=featureSet)
self.styles = self.getParameters(style, [
"typed", "directed", "headsOnly", "graph_kernel", "noAnnType",
"noMasking", "maxFeatures", "genia_limits", "epi_limits",
"id_limits", "rel_limits", "bb_limits", "bi_limits", "co_limits",
"genia_task1", "ontology", "nodalida", "bacteria_renaming",
"trigger_features", "rel_features", "ddi_features", "evex",
"giuliano", "random", "themeOnly", "causeOnly", "no_path",
"entities", "skip_extra_triggers", "headsOnly", "graph_kernel",
"trigger_features", "no_task", "no_dependency",
"disable_entity_features", "disable_terminus_features",
"disable_single_element_features", "disable_ngram_features",
"disable_path_edge_features", "no_linear", "subset", "binary",
"pos_only", "entity_type"
])
if style == None: # no parameters given
style["typed"] = style["directed"] = style["headsOnly"] = True
# self.styles = style
# if "selftrain_group" in self.styles:
# self.selfTrainGroups = set()
# if "selftrain_group-1" in self.styles:
# self.selfTrainGroups.add("-1")
# if "selftrain_group0" in self.styles:
# self.selfTrainGroups.add("0")
# if "selftrain_group1" in self.styles:
# self.selfTrainGroups.add("1")
# if "selftrain_group2" in self.styles:
# self.selfTrainGroups.add("2")
# if "selftrain_group3" in self.styles:
# self.selfTrainGroups.add("3")
# print >> sys.stderr, "Self-train-groups:", self.selfTrainGroups
self.multiEdgeFeatureBuilder = MultiEdgeFeatureBuilder(self.featureSet)
# NOTE Temporarily re-enabling predicted range
#self.multiEdgeFeatureBuilder.definePredictedValueRange([], None)
if self.styles["graph_kernel"]:
from FeatureBuilders.GraphKernelFeatureBuilder import GraphKernelFeatureBuilder
self.graphKernelFeatureBuilder = GraphKernelFeatureBuilder(
self.featureSet)
if self.styles["noAnnType"]:
self.multiEdgeFeatureBuilder.noAnnType = True
if self.styles["noMasking"]:
self.multiEdgeFeatureBuilder.maskNamedEntities = False
if self.styles["maxFeatures"]:
self.multiEdgeFeatureBuilder.maximum = True
if self.styles["genia_task1"]:
self.multiEdgeFeatureBuilder.filterAnnTypes.add("Entity")
self.tokenFeatureBuilder = TokenFeatureBuilder(self.featureSet)
if self.styles["ontology"]:
self.multiEdgeFeatureBuilder.ontologyFeatureBuilder = BioInferOntologyFeatureBuilder(
self.featureSet)
if self.styles["nodalida"]:
self.nodalidaFeatureBuilder = NodalidaFeatureBuilder(
self.featureSet)
if self.styles["bacteria_renaming"]:
self.bacteriaRenamingFeatureBuilder = BacteriaRenamingFeatureBuilder(
self.featureSet)
if self.styles["trigger_features"]:
self.triggerFeatureBuilder = TriggerFeatureBuilder(self.featureSet)
self.triggerFeatureBuilder.useNonNameEntities = True
if self.styles["genia_task1"]:
self.triggerFeatureBuilder.filterAnnTypes.add("Entity")
#self.bioinferOntologies = OntologyUtils.loadOntologies(OntologyUtils.g_bioInferFileName)
if self.styles["rel_features"]:
self.relFeatureBuilder = RELFeatureBuilder(featureSet)
if self.styles["ddi_features"]:
self.drugFeatureBuilder = DrugFeatureBuilder(featureSet)
if self.styles["evex"]:
self.evexFeatureBuilder = EVEXFeatureBuilder(featureSet)
if self.styles["giuliano"]:
self.giulianoFeatureBuilder = GiulianoFeatureBuilder(featureSet)
self.pathLengths = length
assert (self.pathLengths == None)
self.types = types
if self.styles["random"]:
from FeatureBuilders.RandomFeatureBuilder import RandomFeatureBuilder
self.randomFeatureBuilder = RandomFeatureBuilder(self.featureSet)
class MultiEdgeExampleBuilder(ExampleBuilder):
"""
This example builder makes edge examples, i.e. examples describing
the event arguments.
"""
def __init__(self,
style=None,
length=None,
types=[],
featureSet=None,
classSet=None):
if featureSet == None:
featureSet = IdSet()
if classSet == None:
classSet = IdSet(1)
else:
classSet = classSet
assert (classSet.getId("neg") == 1
or (len(classSet.Ids) == 2 and classSet.getId("neg") == -1))
ExampleBuilder.__init__(self, classSet=classSet, featureSet=featureSet)
self.styles = self.getParameters(style, [
"typed", "directed", "headsOnly", "graph_kernel", "noAnnType",
"noMasking", "maxFeatures", "genia_limits", "epi_limits",
"id_limits", "rel_limits", "bb_limits", "bi_limits", "co_limits",
"genia_task1", "ontology", "nodalida", "bacteria_renaming",
"trigger_features", "rel_features", "ddi_features", "evex",
"giuliano", "random", "themeOnly", "causeOnly", "no_path",
"entities", "skip_extra_triggers", "headsOnly", "graph_kernel",
"trigger_features", "no_task", "no_dependency",
"disable_entity_features", "disable_terminus_features",
"disable_single_element_features", "disable_ngram_features",
"disable_path_edge_features", "no_linear", "subset", "binary",
"pos_only", "entity_type"
])
if style == None: # no parameters given
style["typed"] = style["directed"] = style["headsOnly"] = True
# self.styles = style
# if "selftrain_group" in self.styles:
# self.selfTrainGroups = set()
# if "selftrain_group-1" in self.styles:
# self.selfTrainGroups.add("-1")
# if "selftrain_group0" in self.styles:
# self.selfTrainGroups.add("0")
# if "selftrain_group1" in self.styles:
# self.selfTrainGroups.add("1")
# if "selftrain_group2" in self.styles:
# self.selfTrainGroups.add("2")
# if "selftrain_group3" in self.styles:
# self.selfTrainGroups.add("3")
# print >> sys.stderr, "Self-train-groups:", self.selfTrainGroups
self.multiEdgeFeatureBuilder = MultiEdgeFeatureBuilder(self.featureSet)
# NOTE Temporarily re-enabling predicted range
#self.multiEdgeFeatureBuilder.definePredictedValueRange([], None)
if self.styles["graph_kernel"]:
from FeatureBuilders.GraphKernelFeatureBuilder import GraphKernelFeatureBuilder
self.graphKernelFeatureBuilder = GraphKernelFeatureBuilder(
self.featureSet)
if self.styles["noAnnType"]:
self.multiEdgeFeatureBuilder.noAnnType = True
if self.styles["noMasking"]:
self.multiEdgeFeatureBuilder.maskNamedEntities = False
if self.styles["maxFeatures"]:
self.multiEdgeFeatureBuilder.maximum = True
if self.styles["genia_task1"]:
self.multiEdgeFeatureBuilder.filterAnnTypes.add("Entity")
self.tokenFeatureBuilder = TokenFeatureBuilder(self.featureSet)
if self.styles["ontology"]:
self.multiEdgeFeatureBuilder.ontologyFeatureBuilder = BioInferOntologyFeatureBuilder(
self.featureSet)
if self.styles["nodalida"]:
self.nodalidaFeatureBuilder = NodalidaFeatureBuilder(
self.featureSet)
if self.styles["bacteria_renaming"]:
self.bacteriaRenamingFeatureBuilder = BacteriaRenamingFeatureBuilder(
self.featureSet)
if self.styles["trigger_features"]:
self.triggerFeatureBuilder = TriggerFeatureBuilder(self.featureSet)
self.triggerFeatureBuilder.useNonNameEntities = True
if self.styles["genia_task1"]:
self.triggerFeatureBuilder.filterAnnTypes.add("Entity")
#self.bioinferOntologies = OntologyUtils.loadOntologies(OntologyUtils.g_bioInferFileName)
if self.styles["rel_features"]:
self.relFeatureBuilder = RELFeatureBuilder(featureSet)
if self.styles["ddi_features"]:
self.drugFeatureBuilder = DrugFeatureBuilder(featureSet)
if self.styles["evex"]:
self.evexFeatureBuilder = EVEXFeatureBuilder(featureSet)
if self.styles["giuliano"]:
self.giulianoFeatureBuilder = GiulianoFeatureBuilder(featureSet)
self.pathLengths = length
assert (self.pathLengths == None)
self.types = types
if self.styles["random"]:
from FeatureBuilders.RandomFeatureBuilder import RandomFeatureBuilder
self.randomFeatureBuilder = RandomFeatureBuilder(self.featureSet)
def definePredictedValueRange(self, sentences, elementName):
self.multiEdgeFeatureBuilder.definePredictedValueRange(
sentences, elementName)
def getPredictedValueRange(self):
return self.multiEdgeFeatureBuilder.predictedRange
def filterEdgesByType(self, edges, typesToInclude):
if len(typesToInclude) == 0:
return edges
edgesToKeep = []
for edge in edges:
if edge.get("type") in typesToInclude:
edgesToKeep.append(edge)
return edgesToKeep
def getCategoryNameFromTokens(self, sentenceGraph, t1, t2, directed=True):
"""
Example class. Multiple overlapping edges create a merged type.
"""
types = set()
# if sentenceGraph.interactionGraph.has_edge(t1, t2):
# intEdges = sentenceGraph.interactionGraph.get_edge_data(t1, t2, default={})
# # NOTE: Only works if keys are ordered integers
# for i in range(len(intEdges)):
# types.add(intEdges[i]["element"].get("type"))
# if (not directed) and sentenceGraph.interactionGraph.has_edge(t2, t1):
# intEdges = sentenceGraph.interactionGraph.get_edge(t2, t1, default={})
# # NOTE: Only works if keys are ordered integers
# for i in range(len(intEdges)):
# types.add(intEdges[i]["element"].get("type"))
intEdges = sentenceGraph.interactionGraph.getEdges(t1, t2)
if (not directed):
intEdges = intEdges + sentenceGraph.interactionGraph.getEdges(
t2, t1)
for intEdge in intEdges:
types.add(intEdge[2].get("type"))
types = list(types)
types.sort()
categoryName = ""
for name in types:
if categoryName != "":
categoryName += "---"
categoryName += name
if categoryName != "":
return categoryName
else:
return "neg"
def getCategoryName(self,
sentenceGraph,
e1,
e2,
directed=True,
duplicateEntities=None):
"""
Example class. Multiple overlapping edges create a merged type.
"""
# interactions = []
# e1s = [e1]
# if duplicateEntities != None and e1 in duplicateEntities:
# e1s += duplicateEntities[e1]
# e2s = [e2]
# if duplicateEntities != None and e2 in duplicateEntities:
# e2s += duplicateEntities[e2]
# for entity1 in e1s:
# for entity2 in e2s:
# interactions = interactions + sentenceGraph.getInteractions(entity1, entity2)
# if not directed:
# interactions = interactions + sentenceGraph.getInteractions(entity2, entity1)
interactions = sentenceGraph.getInteractions(e1, e2, True)
#print interactions
types = set()
for interaction in interactions:
types.add(interaction[2].get("type"))
types = list(types)
types.sort()
categoryName = ""
for name in types:
if self.styles["causeOnly"] and name != "Cause":
continue
if self.styles["themeOnly"] and name != "Theme":
continue
if categoryName != "":
categoryName += "---"
categoryName += name
if categoryName != "":
return categoryName
else:
return "neg"
def isPotentialRELInteraction(self, e1, e2):
if e1.get("type") == "Protein" and e2.get("type") == "Entity":
return True
else:
return False
def isPotentialBBInteraction(self, e1, e2, sentenceGraph):
#if e1.get("type") == "Bacterium" and e2.get("type") in ["Host", "HostPart", "Geographical", "Environmental", "Food", "Medical", "Soil", "Water"]:
# Note: "Environment" type is misspelled as "Environmental" in the BB-task documentation
if e1.get("type") == "Bacterium" and e2.get("type") in [
"Host", "HostPart", "Geographical", "Environment", "Food",
"Medical", "Soil", "Water"
]:
return True
elif e1.get("type") == "Host" and e2.get("type") == "HostPart":
return True
else:
return False
def getBISuperType(self, eType):
if eType in [
"GeneProduct", "Protein", "ProteinFamily", "PolymeraseComplex"
]:
return "ProteinEntity"
elif eType in [
"Gene", "GeneFamily", "GeneComplex", "Regulon", "Site",
"Promoter"
]:
return "GeneEntity"
else:
return None
def isPotentialBIInteraction(self, e1, e2, sentenceGraph, stats):
e1Type = e1.get("type")
e1SuperType = self.getBISuperType(e1Type)
e2Type = e2.get("type")
e2SuperType = self.getBISuperType(e2Type)
tag = "(" + e1Type + "/" + e2Type + ")"
if e1Type == "Regulon":
if e2SuperType in ["GeneEntity", "ProteinEntity"]:
return True
if e1SuperType == "ProteinEntity":
if e2Type in ["Site", "Promoter", "Gene", "GeneComplex"]:
return True
if e1Type in ["Action", "Transcription", "Expression"]:
return True
if e1Type == "Site":
if e2SuperType == "GeneEntity":
return True
if e1Type == "Promoter":
if e2SuperType in ["GeneEntity", "ProteinEntity"]:
return True
if e1SuperType in ["GeneEntity", "ProteinEntity"]:
if e2SuperType in ["GeneEntity", "ProteinEntity"]:
return True
stats.filter("bi_limits") #+tag)
return False
def isPotentialEPIInteraction(self, e1, e2, sentenceGraph):
if e1.get("type") != "Catalysis":
if e1.get("type") in ["Protein", "Entity"]:
return False
elif e2.get("type") in ["Protein", "Entity"]:
return True
else:
return False
else: # Catalysis
if e2.get("type") != "Entity":
return True
else:
return False
assert False, (e1.get("type"), e2.get("type"))
def isPotentialIDInteraction(self, e1, e2, sentenceGraph):
e1Type = e1.get("type")
e2Type = e2.get("type")
e1IsCore = e1Type in [
"Protein", "Regulon-operon", "Two-component-system", "Chemical",
"Organism"
]
e2IsCore = e2Type in [
"Protein", "Regulon-operon", "Two-component-system", "Chemical",
"Organism"
]
if e1IsCore:
return False
elif e1Type in ["Gene_expression", "Transcription"]:
if e2Type in ["Protein", "Regulon-operon"]:
return True
else:
return False
elif e1Type in ["Protein_catabolism", "Phosphorylation"]:
if e2Type == "Protein":
return True
else:
return False
elif e1Type == "Localization":
if e2IsCore or e2Type == "Entity":
return True
else:
return False
elif e1Type in ["Binding", "Process"]:
if e2IsCore:
return True
else:
return False
elif "egulation" in e1Type:
if e2Type != "Entity":
return True
else:
return False
elif e1Type == "Entity":
if e2IsCore:
return True
else:
return False
assert False, (e1Type, e2Type)
def isPotentialCOInteraction(self, e1, e2, sentenceGraph):
if e1.get("type") == "Exp" and e2.get("type") == "Exp":
anaphoraTok = sentenceGraph.entityHeadTokenByEntity[e1]
antecedentTok = sentenceGraph.entityHeadTokenByEntity[e2]
antecedentTokenFound = False
for token in sentenceGraph.tokens:
if token == antecedentTok:
antecedentTokenFound = True
if token == anaphoraTok: # if, not elif, to take into accoutn cases where e1Tok == e2Tok
if antecedentTokenFound:
return True
else:
return False
assert False
elif e1.get("type") == "Exp" and e2.get("type") == "Protein":
return True
else:
return False
def isPotentialGeniaInteraction(self, e1, e2):
e1Type = e1.get("type")
e2Type = e2.get("type")
if e1Type == "Protein":
return False
elif e1Type in [
"Entity", "Gene_expression", "Transcription",
"Protein_catabolism", "Phosphorylation", "Binding"
]:
if e2Type == "Protein":
return True
else:
return False
elif e1Type == "Localization":
if e2Type in ["Protein", "Entity"]:
return True
else:
return False
elif "egulation" in e1Type:
if e2Type != "Entity":
return True
else:
return False
assert False, (e1Type, e2Type)
def getGoldCategoryName(self,
goldGraph,
entityToGold,
e1,
e2,
directed=True):
if len(entityToGold[e1]) > 0 and len(entityToGold[e2]) > 0:
return self.getCategoryName(goldGraph,
entityToGold[e1][0],
entityToGold[e2][0],
directed=directed)
else:
return "neg"
def buildExamplesFromGraph(self, sentenceGraph, outfile, goldGraph=None):
"""
Build examples for a single sentence. Returns a list of examples.
See Core/ExampleUtils for example format.
"""
#examples = []
exampleIndex = 0
if self.styles["trigger_features"]:
self.triggerFeatureBuilder.initSentence(sentenceGraph)
if self.styles["evex"]:
self.evexFeatureBuilder.initSentence(sentenceGraph)
# Filter entities, if needed
#mergedIds = None
#duplicateEntities = None
#entities = sentenceGraph.entities
#entities, mergedIds, duplicateEntities = self.mergeEntities(sentenceGraph, False) # "no_duplicates" in self.styles)
sentenceGraph.mergeInteractionGraph(True)
entities = sentenceGraph.mergedEntities
entityToDuplicates = sentenceGraph.mergedEntityToDuplicates
self.exampleStats.addValue("Duplicate entities skipped",
len(sentenceGraph.entities) - len(entities))
# Connect to optional gold graph
if goldGraph != None:
entityToGold = EvaluateInteractionXML.mapEntities(
entities, goldGraph.entities)
paths = None
if not self.styles["no_path"]:
##undirected = sentenceGraph.getUndirectedDependencyGraph()
#undirected = self.nxMultiDiGraphToUndirected(sentenceGraph.dependencyGraph)
###undirected = sentenceGraph.dependencyGraph.to_undirected()
####undirected = NX10.MultiGraph(sentenceGraph.dependencyGraph) This didn't work
undirected = sentenceGraph.dependencyGraph.toUndirected()
#paths = NX10.all_pairs_shortest_path(undirected, cutoff=999)
paths = undirected
#for edge in sentenceGraph.dependencyGraph.edges:
# assert edge[2] != None
#for edge in undirected.edges:
# assert edge[2] != None
#if sentenceGraph.sentenceElement.get("id") == "GENIA.d70.s5":
# print [(x[0].get("id"), x[1].get("id"), x[2].get("id")) for x in sentenceGraph.dependencyGraph.edges]
# Generate examples based on interactions between entities or interactions between tokens
if self.styles["entities"]:
loopRange = len(entities)
else:
loopRange = len(sentenceGraph.tokens)
for i in range(loopRange - 1):
for j in range(i + 1, loopRange):
eI = None
eJ = None
if self.styles["entities"]:
eI = entities[i]
eJ = entities[j]
tI = sentenceGraph.entityHeadTokenByEntity[eI]
tJ = sentenceGraph.entityHeadTokenByEntity[eJ]
#if "no_ne_interactions" in self.styles and eI.get("isName") == "True" and eJ.get("isName") == "True":
# continue
if eI.get("type") == "neg" or eJ.get("type") == "neg":
continue
if self.styles["skip_extra_triggers"]:
if eI.get("source") != None or eJ.get(
"source") != None:
continue
else:
tI = sentenceGraph.tokens[i]
tJ = sentenceGraph.tokens[j]
# only consider paths between entities (NOTE! entities, not only named entities)
if self.styles["headsOnly"]:
if (len(sentenceGraph.tokenIsEntityHead[tI]) == 0) or (len(
sentenceGraph.tokenIsEntityHead[tJ]) == 0):
continue
if self.styles["directed"]:
# define forward
if self.styles["entities"]:
categoryName = self.getCategoryName(
sentenceGraph, eI, eJ, True)
if goldGraph != None:
categoryName = self.getGoldCategoryName(
goldGraph, entityToGold, eI, eJ, True)
else:
categoryName = self.getCategoryNameFromTokens(
sentenceGraph, tI, tJ, True)
# make forward
self.exampleStats.beginExample(categoryName)
makeExample = True
if self.styles[
"genia_limits"] and not self.isPotentialGeniaInteraction(
eI, eJ):
makeExample = False
self.exampleStats.filter("genia_limits")
if self.styles["genia_task1"] and (
eI.get("type") == "Entity"
or eJ.get("type") == "Entity"):
makeExample = False
self.exampleStats.filter("genia_task1")
if self.styles[
"rel_limits"] and not self.isPotentialRELInteraction(
eI, eJ):
makeExample = False
self.exampleStats.filter("rel_limits")
if self.styles[
"co_limits"] and not self.isPotentialCOInteraction(
eI, eJ, sentenceGraph):
makeExample = False
self.exampleStats.filter("co_limits")
if self.styles[
"bb_limits"] and not self.isPotentialBBInteraction(
eI, eJ, sentenceGraph):
makeExample = False
self.exampleStats.filter("bb_limits")
if categoryName != "neg":
self.exampleStats.filter("bb_limits(" +
categoryName + ":" +
eI.get("type") + "/" +
eJ.get("type") + ")")
if self.styles[
"bi_limits"] and not self.isPotentialBIInteraction(
eI, eJ, sentenceGraph, self.exampleStats):
makeExample = False
#self.exampleStats.filter("bi_limits")
if self.styles[
"epi_limits"] and not self.isPotentialEPIInteraction(
eI, eJ, sentenceGraph):
makeExample = False
self.exampleStats.filter("epi_limits")
if self.styles[
"id_limits"] and not self.isPotentialIDInteraction(
eI, eJ, sentenceGraph):
makeExample = False
self.exampleStats.filter("id_limits")
# if self.styles["selftrain_limits"] and (eI.get("selftrain") == "False" or eJ.get("selftrain") == "False"):
# makeExample = False
# self.exampleStats.filter("selftrain_limits")
# if self.styles["selftrain_group"] and (eI.get("selftraingroup") not in self.selfTrainGroups or eJ.get("selftraingroup") not in self.selfTrainGroups):
# makeExample = False
# self.exampleStats.filter("selftrain_group")
if self.styles["pos_only"] and categoryName == "neg":
makeExample = False
self.exampleStats.filter("pos_only")
if makeExample:
#examples.append( self.buildExample(tI, tJ, paths, sentenceGraph, categoryName, exampleIndex, eI, eJ) )
ExampleUtils.appendExamples([
self.buildExample(tI, tJ, paths, sentenceGraph,
categoryName, exampleIndex, eI,
eJ)
], outfile)
exampleIndex += 1
self.exampleStats.endExample()
# define reverse
if self.styles["entities"]:
categoryName = self.getCategoryName(
sentenceGraph, eJ, eI, True)
if goldGraph != None:
categoryName = self.getGoldCategoryName(
goldGraph, entityToGold, eJ, eI, True)
else:
categoryName = self.getCategoryNameFromTokens(
sentenceGraph, tJ, tI, True)
# make reverse
self.exampleStats.beginExample(categoryName)
makeExample = True
if self.styles[
"genia_limits"] and not self.isPotentialGeniaInteraction(
eJ, eI):
makeExample = False
self.exampleStats.filter("genia_limits")
if self.styles["genia_task1"] and (
eI.get("type") == "Entity"
or eJ.get("type") == "Entity"):
makeExample = False
self.exampleStats.filter("genia_task1")
if self.styles[
"rel_limits"] and not self.isPotentialRELInteraction(
eJ, eI):
makeExample = False
self.exampleStats.filter("rel_limits")
if self.styles[
"co_limits"] and not self.isPotentialCOInteraction(
eJ, eI, sentenceGraph):
makeExample = False
self.exampleStats.filter("co_limits")
if self.styles[
"bb_limits"] and not self.isPotentialBBInteraction(
eJ, eI, sentenceGraph):
makeExample = False
self.exampleStats.filter("bb_limits")
if categoryName != "neg":
self.exampleStats.filter("bb_limits(" +
categoryName + ":" +
eJ.get("type") + "/" +
eI.get("type") + ")")
if self.styles[
"bi_limits"] and not self.isPotentialBIInteraction(
eJ, eI, sentenceGraph, self.exampleStats):
makeExample = False
#self.exampleStats.filter("bi_limits")
if self.styles[
"epi_limits"] and not self.isPotentialEPIInteraction(
eJ, eI, sentenceGraph):
makeExample = False
self.exampleStats.filter("epi_limits")
if self.styles[
"id_limits"] and not self.isPotentialIDInteraction(
eJ, eI, sentenceGraph):
makeExample = False
self.exampleStats.filter("id_limits")
# if self.styles["selftrain_limits"] and (eI.get("selftrain") == "False" or eJ.get("selftrain") == "False"):
# makeExample = False
# self.exampleStats.filter("selftrain_limits")
# if self.styles["selftrain_group"] and (eI.get("selftraingroup") not in self.selfTrainGroups or eJ.get("selftraingroup") not in self.selfTrainGroups):
# makeExample = False
# self.exampleStats.filter("selftrain_group")
if self.styles["pos_only"] and categoryName == "neg":
makeExample = False
self.exampleStats.filter("pos_only")
if makeExample:
#examples.append( self.buildExample(tJ, tI, paths, sentenceGraph, categoryName, exampleIndex, eJ, eI) )
ExampleUtils.appendExamples([
self.buildExample(tJ, tI, paths, sentenceGraph,
categoryName, exampleIndex, eJ,
eI)
], outfile)
exampleIndex += 1
self.exampleStats.endExample()
else:
if self.styles["entities"]:
categoryName = self.getCategoryName(
sentenceGraph, eI, eJ, False)
else:
categoryName = self.getCategoryNameFromTokens(
sentenceGraph, tI, tJ, False)
self.exampleStats.beginExample(categoryName)
forwardExample = self.buildExample(tI, tJ, paths,
sentenceGraph,
categoryName,
exampleIndex, eI, eJ)
if not self.styles["graph_kernel"]:
reverseExample = self.buildExample(
tJ, tI, paths, sentenceGraph, categoryName,
exampleIndex, eJ, eI)
forwardExample[2].update(reverseExample[2])
#examples.append(forwardExample)
ExampleUtils.appendExamples([forwardExample], outfile)
exampleIndex += 1
self.exampleStats.endExample()
#return examples
return exampleIndex
def buildExample(self,
token1,
token2,
paths,
sentenceGraph,
categoryName,
exampleIndex,
entity1=None,
entity2=None):
"""
Build a single directed example for the potential edge between token1 and token2
"""
# dummy return for speed testing
#return (sentenceGraph.getSentenceId()+".x"+str(exampleIndex),1,{},{})
# define features
features = {}
if True: #token1 != token2 and paths.has_key(token1) and paths[token1].has_key(token2):
#if token1 != token2 and paths.has_key(token1) and paths[token1].has_key(token2):
# path = paths[token1][token2]
#else:
# path = [token1, token2]
if not self.styles["no_path"]:
# directedPath reduces performance by 0.01 pp
#directedPath = sentenceGraph.dependencyGraph.getPaths(token1, token2)
#if len(directedPath) == 0:
# directedPath = sentenceGraph.dependencyGraph.getPaths(token2, token1)
# for dp in directedPath:
# dp.reverse()
#if len(directedPath) == 0:
# path = paths.getPaths(token1, token2)
#else:
# path = directedPath
path = paths.getPaths(token1, token2)
if len(path) > 0:
#if len(path) > 1:
# print len(path)
path = path[0]
pathExists = True
else:
path = [token1, token2]
pathExists = False
else:
path = [token1, token2]
pathExists = False
#print token1.get("id"), token2.get("id")
assert (self.pathLengths == None)
if self.pathLengths == None or len(path) - 1 in self.pathLengths:
# if not "no_ontology" in self.styles:
# self.ontologyFeatureBuilder.setFeatureVector(features)
# self.ontologyFeatureBuilder.buildOntologyFeaturesForPath(sentenceGraph, path)
# self.ontologyFeatureBuilder.setFeatureVector(None)
if self.styles["trigger_features"]: # F 85.52 -> 85.55
self.triggerFeatureBuilder.setFeatureVector(features)
self.triggerFeatureBuilder.tag = "trg1_"
self.triggerFeatureBuilder.buildFeatures(token1)
self.triggerFeatureBuilder.tag = "trg2_"
self.triggerFeatureBuilder.buildFeatures(token2)
self.triggerFeatureBuilder.setFeatureVector(None)
# REL features
if self.styles["rel_features"] and not self.styles["no_task"]:
self.relFeatureBuilder.setFeatureVector(features)
self.relFeatureBuilder.tag = "rel1_"
self.relFeatureBuilder.buildAllFeatures(
sentenceGraph.tokens,
sentenceGraph.tokens.index(token1))
self.relFeatureBuilder.tag = "rel2_"
self.relFeatureBuilder.buildAllFeatures(
sentenceGraph.tokens,
sentenceGraph.tokens.index(token2))
self.relFeatureBuilder.setFeatureVector(None)
if self.styles[
"bacteria_renaming"] and not self.styles["no_task"]:
self.bacteriaRenamingFeatureBuilder.setFeatureVector(
features)
self.bacteriaRenamingFeatureBuilder.buildPairFeatures(
entity1, entity2)
#self.bacteriaRenamingFeatureBuilder.buildSubstringFeatures(entity1, entity2) # decreases perf. 74.76 -> 72.41
self.bacteriaRenamingFeatureBuilder.setFeatureVector(None)
if self.styles["co_limits"] and not self.styles["no_task"]:
e1Offset = Range.charOffsetToSingleTuple(
entity1.get("charOffset"))
e2Offset = Range.charOffsetToSingleTuple(
entity2.get("charOffset"))
if Range.contains(e1Offset, e2Offset):
features[self.featureSet.getId("e1_contains_e2")] = 1
if entity2.get("isName") == "True":
features[self.featureSet.getId(
"e1_contains_e2name")] = 1
if Range.contains(e2Offset, e1Offset):
features[self.featureSet.getId("e2_contains_e1")] = 1
if entity1.get("isName") == "True":
features[self.featureSet.getId(
"e2_contains_e1name")] = 1
if self.styles["ddi_features"]:
self.drugFeatureBuilder.setFeatureVector(features)
self.drugFeatureBuilder.tag = "ddi_"
self.drugFeatureBuilder.buildPairFeatures(entity1, entity2)
if self.styles["ddi_mtmx"]:
self.drugFeatureBuilder.buildMTMXFeatures(
entity1, entity2)
self.drugFeatureBuilder.setFeatureVector(None)
#if "graph_kernel" in self.styles or not "no_dependency" in self.styles:
# #print "Getting edges"
# if token1 != token2 and pathExists:
# #print "g1"
# edges = self.multiEdgeFeatureBuilder.getEdges(sentenceGraph.dependencyGraph, path)
# #print "g2"
# else:
# edges = None
if self.styles["graph_kernel"]:
self.graphKernelFeatureBuilder.setFeatureVector(
features, entity1, entity2)
self.graphKernelFeatureBuilder.buildGraphKernelFeatures(
sentenceGraph, path)
self.graphKernelFeatureBuilder.setFeatureVector(None)
if self.styles["entity_type"]:
features[self.featureSet.getId("e1_" +
entity1.get("type"))] = 1
features[self.featureSet.getId("e2_" +
entity2.get("type"))] = 1
features[self.featureSet.getId("distance_" +
str(len(path)))] = 1
if not self.styles["no_dependency"]:
#print "Dep features"
self.multiEdgeFeatureBuilder.setFeatureVector(
features, entity1, entity2)
#self.multiEdgeFeatureBuilder.buildStructureFeatures(sentenceGraph, paths) # remove for fast
if not self.styles["disable_entity_features"]:
self.multiEdgeFeatureBuilder.buildEntityFeatures(
sentenceGraph)
self.multiEdgeFeatureBuilder.buildPathLengthFeatures(path)
if not self.styles["disable_terminus_features"]:
self.multiEdgeFeatureBuilder.buildTerminusTokenFeatures(
path, sentenceGraph) # remove for fast
if not self.styles["disable_single_element_features"]:
self.multiEdgeFeatureBuilder.buildSingleElementFeatures(
path, sentenceGraph)
if not self.styles["disable_ngram_features"]:
#print "NGrams"
self.multiEdgeFeatureBuilder.buildPathGrams(
2, path, sentenceGraph) # remove for fast
self.multiEdgeFeatureBuilder.buildPathGrams(
3, path, sentenceGraph) # remove for fast
self.multiEdgeFeatureBuilder.buildPathGrams(
4, path, sentenceGraph) # remove for fast
#self.buildEdgeCombinations(path, edges, sentenceGraph, features) # remove for fast
#if edges != None:
# self.multiEdgeFeatureBuilder.buildTerminusFeatures(path[0], edges[0][1]+edges[1][0], "t1", sentenceGraph) # remove for fast
# self.multiEdgeFeatureBuilder.buildTerminusFeatures(path[-1], edges[len(path)-1][len(path)-2]+edges[len(path)-2][len(path)-1], "t2", sentenceGraph) # remove for fast
if not self.styles["disable_path_edge_features"]:
self.multiEdgeFeatureBuilder.buildPathEdgeFeatures(
path, sentenceGraph)
self.multiEdgeFeatureBuilder.buildSentenceFeatures(
sentenceGraph)
self.multiEdgeFeatureBuilder.setFeatureVector(None)
if self.styles["nodalida"]:
self.nodalidaFeatureBuilder.setFeatureVector(
features, entity1, entity2)
shortestPaths = self.nodalidaFeatureBuilder.buildShortestPaths(
sentenceGraph.dependencyGraph, path)
print shortestPaths
if len(shortestPaths) > 0:
self.nodalidaFeatureBuilder.buildNGrams(
shortestPaths, sentenceGraph)
self.nodalidaFeatureBuilder.setFeatureVector(None)
if not self.styles["no_linear"]:
self.tokenFeatureBuilder.setFeatureVector(features)
for i in range(len(sentenceGraph.tokens)):
if sentenceGraph.tokens[i] == token1:
token1Index = i
if sentenceGraph.tokens[i] == token2:
token2Index = i
linearPreTag = "linfw_"
if token1Index > token2Index:
token1Index, token2Index = token2Index, token1Index
linearPreTag = "linrv_"
self.tokenFeatureBuilder.buildLinearOrderFeatures(
token1Index, sentenceGraph, 2, 2, preTag="linTok1")
self.tokenFeatureBuilder.buildLinearOrderFeatures(
token2Index, sentenceGraph, 2, 2, preTag="linTok2")
# Before, middle, after
# self.tokenFeatureBuilder.buildTokenGrams(0, token1Index-1, sentenceGraph, "bf")
# self.tokenFeatureBuilder.buildTokenGrams(token1Index+1, token2Index-1, sentenceGraph, "bw")
# self.tokenFeatureBuilder.buildTokenGrams(token2Index+1, len(sentenceGraph.tokens)-1, sentenceGraph, "af")
# before-middle, middle, middle-after
# self.tokenFeatureBuilder.buildTokenGrams(0, token2Index-1, sentenceGraph, linearPreTag+"bf", max=2)
# self.tokenFeatureBuilder.buildTokenGrams(token1Index+1, token2Index-1, sentenceGraph, linearPreTag+"bw", max=2)
# self.tokenFeatureBuilder.buildTokenGrams(token1Index+1, len(sentenceGraph.tokens)-1, sentenceGraph, linearPreTag+"af", max=2)
self.tokenFeatureBuilder.setFeatureVector(None)
if self.styles["random"]:
self.randomFeatureBuilder.setFeatureVector(features)
self.randomFeatureBuilder.buildRandomFeatures(100, 0.01)
self.randomFeatureBuilder.setFeatureVector(None)
if self.styles["genia_limits"] and not self.styles["no_task"]:
e1Type = entity1.get("type")
e2Type = entity2.get("type")
assert (entity1.get("isName") == "False")
if entity2.get("isName") == "True":
features[self.featureSet.getId(
"GENIA_target_protein")] = 1
else:
features[self.featureSet.getId(
"GENIA_nested_event")] = 1
if e1Type.find(
"egulation"
) != -1: # leave r out to avoid problems with capitalization
if entity2.get("isName") == "True":
features[self.featureSet.getId(
"GENIA_regulation_of_protein")] = 1
else:
features[self.featureSet.getId(
"GENIA_regulation_of_event")] = 1
if self.styles["bi_limits"]:
# Make features based on entity types
e1Type = entity1.get("type")
e2Type = entity2.get("type")
e1SuperType = str(self.getBISuperType(e1Type))
e2SuperType = str(self.getBISuperType(e2Type))
features[self.featureSet.getId("BI_e1_" + e1Type)] = 1
features[self.featureSet.getId("BI_e2_" + e2Type)] = 1
features[self.featureSet.getId("BI_e1sup_" +
e1SuperType)] = 1
features[self.featureSet.getId("BI_e2sup_" +
e2SuperType)] = 1
features[self.featureSet.getId("BI_e1e2_" + e1Type + "_" +
e2Type)] = 1
features[self.featureSet.getId("BI_e1e2sup_" +
e1SuperType + "_" +
e2SuperType)] = 1
if self.styles["evex"]:
self.evexFeatureBuilder.setFeatureVector(
features, entity1, entity2)
self.evexFeatureBuilder.buildEdgeFeatures(
entity1, entity2, token1, token2, path, sentenceGraph)
self.evexFeatureBuilder.setFeatureVector(None)
if self.styles["giuliano"]:
self.giulianoFeatureBuilder.setFeatureVector(
features, entity1, entity2)
self.giulianoFeatureBuilder.buildEdgeFeatures(
entity1, entity2, token1, token2, path, sentenceGraph)
self.giulianoFeatureBuilder.setFeatureVector(None)
else:
features[self.featureSet.getId("always_negative")] = 1
if self.styles["subset"]:
features[self.featureSet.getId("out_of_scope")] = 1
else:
features[self.featureSet.getId("always_negative")] = 1
if self.styles["subset"]:
features[self.featureSet.getId("out_of_scope")] = 1
path = [token1, token2]
# define extra attributes
#if int(path[0].get("id").split("_")[-1]) < int(path[-1].get("id").split("_")[-1]):
if int(path[0].get("charOffset").split("-")[0]) < int(
path[-1].get("charOffset").split("-")[0]):
#extra = {"xtype":"edge","type":"i","t1":path[0],"t2":path[-1]}
extra = {
"xtype": "edge",
"type": "i",
"t1": path[0].get("id"),
"t2": path[-1].get("id")
}
extra["deprev"] = False
else:
#extra = {"xtype":"edge","type":"i","t1":path[-1],"t2":path[0]}
extra = {
"xtype": "edge",
"type": "i",
"t1": path[-1].get("id"),
"t2": path[0].get("id")
}
extra["deprev"] = True
if entity1 != None:
#extra["e1"] = entity1
extra["e1"] = entity1.get("id")
if sentenceGraph.mergedEntityToDuplicates != None:
#extra["e1GoldIds"] = mergedEntityIds[entity1]
extra["e1DuplicateIds"] = ",".join([
x.get("id")
for x in sentenceGraph.mergedEntityToDuplicates[entity1]
])
if entity2 != None:
#extra["e2"] = entity2
extra["e2"] = entity2.get("id")
if sentenceGraph.mergedEntityToDuplicates != None:
extra["e2DuplicateIds"] = ",".join([
x.get("id")
for x in sentenceGraph.mergedEntityToDuplicates[entity2]
])
#extra["e2GoldIds"] = mergedEntityIds[entity2]
extra["categoryName"] = categoryName
if self.styles["bacteria_renaming"]:
if entity1.get("text") != None and entity1.get("text") != "":
extra["e1t"] = entity1.get("text").replace(" ", "---").replace(
":", "-COL-")
if entity2.get("text") != None and entity2.get("text") != "":
extra["e2t"] = entity2.get("text").replace(" ", "---").replace(
":", "-COL-")
sentenceOrigId = sentenceGraph.sentenceElement.get("origId")
if sentenceOrigId != None:
extra["SOID"] = sentenceOrigId
# make example
if self.styles["binary"]:
if categoryName != "neg":
category = 1
else:
category = -1
categoryName = "i"
else:
category = self.classSet.getId(categoryName)
# NOTE: temporarily disable for replicating 110310 experiment
#features[self.featureSet.getId("extra_constant")] = 1
return (sentenceGraph.getSentenceId() + ".x" + str(exampleIndex),
category, features, extra)
