Interface | Description |
---|---|
CacheKeyGenerator | |
CorpusKernelEvaluator | |
Kernel |
Class | Description |
---|---|
AttributeProductKernel |
Expects numeric values as input.
|
CacheKernel |
consolidate caching in this class.
|
ConvolutionKernel |
Apply the delegate kernel to the children of the given nodes.
|
CorpusKernelEvaluatorImpl | |
EqualityKernel | |
LinKernel | |
MethodCachingInterceptor |
simple caching interceptor.
|
NamedEntityNegationKernel |
Evaluate negation status and certainty of named entities.
|
NodeAttributeKernel |
Extract a node attribute and run the delegate kernel on the attribute.
|
NormKernel |
Return norm of delegate kernel:
k(x,y)/sqrt(k(x,x)*k(y,y) . |
ProductKernel |
compute the product of delegate kernels
|
SemanticSimKernel | |
SemanticTypeKernel |
Before comparing semantic distance, use this kernel to filter by semantic
type.
|
SumKernel |
apply all the delegate kernels to the objects, sum them up
|
SymmetricPairCacheKeyGenerator |
cache key for a method that takes 2 arguments, and is symmetric - the order
of the arguments doesn't matter.
|
TreePrinter | |
WeightedPolynomialMixingKernel |
weighted polynomial mixing kernel:
(\sum w_i * k(x_i, y_i) \div \sum w_i ) ^ l
Apply the delegate kernel to the respective 'parts' of this node (x_i,
y_i)
Multiply the result by the weight (w_i * k(x_i,y_i))
Sum everything up, divide by the sum of the weights
take the power of everything to l
WeightedPolynomialMixingKernel.pow the power to raise things to
WeightedPolynomialMixingKernel.attributeKey the children of this node form a 'vector'. |