|
|
|
|
|
| buildGraphKernelFeatures(self,
sentenceGraph,
path) |
source code
|
|
|
|
_matrixToFeatures(self,
W,
labels)
Linearizes the representation of the graph |
source code
|
|
|
|
| _prepareMatrix(self,
adjacencyMatrix,
node_count,
dtyp=<type 'numpy.float64'>) |
source code
|
|
|
|
_getTokenId(self,
tokenElement)
Returns the position id of the token |
source code
|
|
|
|
| _getTokenText(self,
path,
sentenceGraph,
token) |
source code
|
|
|
|
_buildAdjacencyMatrix(self,
sentenceGraph,
path,
edges,
floattype=<type 'numpy.float64'>,
directed=True,
linearOrderWeight=0.9)
Returns a Numpy-matrix |
source code
|
|
|
|
_setAllDependencyWeights(self,
sentenceGraph,
weights,
weight)
All weights are set to the given value |
source code
|
|
|
|
|
|
|
_setDependencyWeightsByPath(self,
edges,
weights,
weight)
The weights of all dependencies in specified paths are set to the
given value |
source code
|
|
|
|
_reduceWeightByDistance(self,
sentenceGraph,
weights,
zeroDistanceThreshold=0.9,
reduceFactor=0.5)
Reduces the weight of dependencies based on their distance from the
nearest dependency whose weight is >= the threshold. |
source code
|
|
|
|
_addPositionTags(self,
sentenceGraph,
entity1Tokens,
entity2Tokens)
Sets a prefix to the tokens ppiText based on their linear order in
the sentence. |
source code
|
|
|
Inherited from FeatureBuilder.FeatureBuilder:
getEntityType,
getPOSSuperType,
getTokenAnnotatedType,
getTokenFeatures,
normalizeFeatureVector,
setFeature,
setFeatureVector,
setTag
|