Skip to content
Snippets Groups Projects
Commit 8174cec5 authored by Julien Breton's avatar Julien Breton
Browse files

add marker augmentation

parent 97d4baab
No related branches found
No related tags found
No related merge requests found
......@@ -33,3 +33,6 @@ class MarkersList:
def getReasonMarkers(self):
return ["visé", "visée", "visées", "prévues", "prévue", "servant", "afin de", "en vue de", "en vue d'",
"à effet"]
def getArtifactMarkers(self):
return ["équipements", "embrayage", "taxe", "taxes", "avertissement", "délai", "delais", "autorisations", "cours"]
This diff is collapsed.
from modules.evaluators.recall_precision_f1_f2 import recall_precision_f1_f2
from modules.marker_enhancer.s1 import MarkerEnhancerSemanticRessources
from data.MarkersList import MarkersList
top_setting = 10
def main():
markerList = {
"location": MarkersList().getLocationMarkers(),
"condition": MarkersList().getConditionMarkers(),
"exception": MarkersList().getExceptionMarkers(),
"actor": MarkersList().getActorMarkers(),
"modality": MarkersList().getModalityMarkers(),
"time": MarkersList().getTimeMarkers(),
"artifact": MarkersList().getArtifactMarkers(),
}
for key, value in markerList.items():
mkhsr = MarkerEnhancerSemanticRessources(value).exec(lemma_setting=True, top_setting=top_setting)
fn = len(mkhsr.getFullMarkers() - mkhsr.getAugmentedMarkersWithoutDoublon())
tp = len(mkhsr.getFullMarkers() & mkhsr.getAugmentedMarkersWithoutDoublon())
fp = len(mkhsr.getAugmentedMarkersWithoutDoublon() - (mkhsr.getFullMarkers() & mkhsr.getAugmentedMarkersWithoutDoublon()))
print(key, recall_precision_f1_f2(fn, tp, fp))
if __name__ == '__main__':
main()
from modules.evaluators.recall_precision_f1_f2 import recall_precision_f1_f2
from modules.marker_enhancer.s2 import MarkerEnhancerLegalRoberta
from data.MarkersList import MarkersList
top_setting = 10
def main():
markerList = {
"location": MarkersList().getLocationMarkers(),
"condition": MarkersList().getConditionMarkers(),
"exception": MarkersList().getExceptionMarkers(),
"actor": MarkersList().getActorMarkers(),
"modality": MarkersList().getModalityMarkers(),
"time": MarkersList().getTimeMarkers(),
"artifact": MarkersList().getArtifactMarkers(),
}
for key, value in markerList.items():
mkhlr = MarkerEnhancerLegalRoberta(value, ['./data/evalQS.csv'], key).exec()
fn = len(mkhsr.getFullMarkers() - mkhlr.getAugmentedMarkersWithoutDoublon())
tp = len(mkhlr.getFullMarkers() & mkhlr.getAugmentedMarkersWithoutDoublon())
fp = len(mkhlr.getAugmentedMarkersWithoutDoublon() - (
mkhlr.getFullMarkers() & mkhlr.getAugmentedMarkersWithoutDoublon()))
print(key, recall_precision_f1_f2(fn, tp, fp))
if __name__ == '__main__':
main()
......@@ -66,7 +66,7 @@ class MarkerEnhancerSemanticRessources:
return self.augmentedMarkers
def getFullMarkers(self):
return self.fullMarkers
return set(self.fullMarkers)
def getOtherLemma(self, lemma, df):
resultats = df.loc[df[0] == lemma]
......@@ -84,3 +84,9 @@ class MarkerEnhancerSemanticRessources:
mySet2.remove(lemma)
return list(mySet2)
def getAugmentedMarkersWithoutDoublon(self):
augmentedMarkersWithoutDoublon = set()
for key, value in self.augmentedMarkers.items():
augmentedMarkersWithoutDoublon.update(list(value))
return augmentedMarkersWithoutDoublon
......@@ -87,4 +87,10 @@ class MarkerEnhancerLegalRoberta:
return self.augmentedMarkers
def getFullMarkers(self):
return self.fullMarkers
\ No newline at end of file
return self.fullMarkers
def getAugmentedMarkersWithoutDoublon(self):
augmentedMarkersWithoutDoublon = set()
for key, value in self.augmentedMarkers.items():
augmentedMarkersWithoutDoublon.update(list(value))
return augmentedMarkersWithoutDoublon
\ No newline at end of file
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment