| Interface | Description |
|---|---|
| Aligner<LeftT,RightT> |
Something which can produce an
Alignment. |
| Alignment<LeftT,RightT> |
Represents a 'pairing up' of two groups of objects of the same type, the
key set and the
test set. |
| BootstrapInspector.BootstrapStrategy<ObsT,SummaryT> | |
| BootstrapInspector.ObservationSummarizer<ObsT,SummaryT> |
A strategy for reducing a pair observation to a some sort of summary we can later bootstrap
sample over.
|
| BootstrapInspector.SummaryAggregator<SummaryT> |
A strategy of aggregating bootstrap samples into some sort of final output.
|
| ProvenancedAligner<LeftT,LeftProvT,RightT,RightProvT> |
Something which can produce a
ProvenancedAlignment. |
| ProvenancedAlignment<LeftT,LeftProvT,RightT,RightProvT> |
An
Alignment which also tracks "provenances" (which can be arbitrary metadata) for the
items being aligned. |
| ScoringEventObserver<KeyT,TestT> |
An object which records scoring events of objects where the expected (gold) item is of type
KeyT and the
predicted (system) item is of type TestT. |
| Class | Description |
|---|---|
| AggregateBinaryFScoresInspector<KeyT,TestT> |
Observes an
Alignment between system items on the left and gold standard items on the
right. |
| Alignments |
Utility methods for dealing with
Alignments. |
| BinaryErrorLogger<KeyT extends HasDocID,TestT extends HasDocID> |
An inspector which logs to a file the binary confusion matrix for an alignment.
|
| BinaryFScoreBootstrapStrategy<T> |
A bootstrap sampling strategy for precision, recall, F-score, and accuracy.
|
| BootstrapInspector<ObsT,SummaryT> |
An inspector which can use a supplied strategy to do some sort of scoring or inspection of a
corpus using bootstrapping.
|
| BootstrapInspector.Builder<ObsT,SummaryT> | |
| BrokenDownFMeasureAggregator |
If we have some metric broken down by categories which results in F-measure counts, these
will print bootstrapped precision/recall/F-measure scores.
|
| BrokenDownPRFAggregator |
If we have some metric broken down by categories which results in confusion matrices, these
will print bootstrapped precision/recall/F-measure scores.
|
| EquivalenceBasedProvenancedAligner<LeftT,RightT,EqClassT> |
Creates
ProvenancedAlignments based on grouping items into equivalence classes. |
| EquivalenceBasedProvenancedAlignment<EqClassT,LeftT,RightT> |
A
ProvenancedAlignment based on grouping items into equivalence classes. |
| EvalPair<KeyT,TestT> |
A pair of things being compared to each other, where one is in some sense "the correct one" (the
key).
|
| InspectionNode<InT> |
An inspection tree node which inspects things.
|
| InspectorTreeDSL |
Inspector trees are a convenient way to write scorers (and possibly other corpus observers) which
compute several related things from one input.
|
| InspectorTreeDSL.InspectionBuilder<InT> | |
| InspectorTreeNode<OutT> |
A node in an inspector tree.
|
| MultimapAlignment<LeftT,RightT> |
A default implementation for
Alignment. |
| MultimapAlignment.Builder<LeftT,RightT> | |
| TypeConfusionInspector<LeftRightT> |
Observes an
Alignment between system items on the left and gold standard items on the
right and creates a confusion matrix between them based on a specified confusion equivalence
and labelling. |
Copyright © 2016. All rights reserved.