edu.stanford.nlp.naturalli

Interface ClauseSplitter

All Superinterfaces:

java.util.function.BiFunction<SemanticGraph,java.lang.Boolean,ClauseSplitterSearchProblem>

public interface ClauseSplitter
extends java.util.function.BiFunction<SemanticGraph,java.lang.Boolean,ClauseSplitterSearchProblem>

Just a convenience alias for a clause splitting search problem factory. Mostly here to form a nice parallel with ForwardEntailer.

Author:

Gabor Angeli

Nested Class Summary

Nested Classes

Modifier and Type	Interface and Description
static class	ClauseSplitter.ClauseClassifierLabel

Field Summary

Fields

Modifier and Type	Field and Description
static Redwood.RedwoodChannels	log A logger for this class

Method Summary

Modifier and Type	Method and Description
static ClauseSplitter	load(java.lang.String serializedModel) Load a factory model from a given path.
static ClauseSplitter	train(java.util.stream.Stream<Pair<CoreMap,java.util.Collection<Pair<Span,Span>>>> trainingData, java.io.File modelPath, java.io.File trainingDataDump)
static ClauseSplitter	train(java.util.stream.Stream<Pair<CoreMap,java.util.Collection<Pair<Span,Span>>>> trainingData, java.util.Optional<java.io.File> modelPath, java.util.Optional<java.io.File> trainingDataDump, ClauseSplitterSearchProblem.Featurizer featurizer) Train a clause searcher factory.

Methods inherited from interface java.util.function.BiFunction

andThen, apply

Field Detail

log

static final Redwood.RedwoodChannels log

A logger for this class

Method Detail

train

static ClauseSplitter train(java.util.stream.Stream<Pair<CoreMap,java.util.Collection<Pair<Span,Span>>>> trainingData,
                            java.util.Optional<java.io.File> modelPath,
                            java.util.Optional<java.io.File> trainingDataDump,
                            ClauseSplitterSearchProblem.Featurizer featurizer)

Train a clause searcher factory. That is, train a classifier for which arcs should be new clauses.

Parameters:

trainingData - The training data. This is a stream of triples of:

1. The sentence containing a known extraction.
2. The span of the subject in the sentence, as a token span.
3. The span of the object in the sentence, as a token span.

modelPath - The path to save the model to. This is useful for load(String).

trainingDataDump - The path to save the training data, as a set of labeled featurized datums.

featurizer - The featurizer to use for this classifier.

Returns:

A factory for creating searchers from a given dependency tree.

train

static ClauseSplitter train(java.util.stream.Stream<Pair<CoreMap,java.util.Collection<Pair<Span,Span>>>> trainingData,
                            java.io.File modelPath,
                            java.io.File trainingDataDump)

load

static ClauseSplitter load(java.lang.String serializedModel)
                    throws java.io.IOException

Load a factory model from a given path. This can be trained with ClauseSplitter#train(Stream, Optional, Optional, Featurizer).

Returns:

A function taking a dependency tree, and returning a clause searcher.

Throws:

java.io.IOException