Class DistributionalLexiconGenerator<T extends Tag>
- java.lang.Object
-
- com.gengoai.hermes.lexicon.generation.DistributionalLexiconGenerator<T>
-
- Type Parameters:
T
- the tag type parameter
- All Implemented Interfaces:
LexiconGenerator<T>
public class DistributionalLexiconGenerator<T extends Tag> extends Object implements LexiconGenerator<T>
Generates a lexicon based on similarity in an embedding space where positive and negative examples can be given per tag category.- Author:
- David B. Bracewell
-
-
Constructor Summary
Constructors Constructor Description DistributionalLexiconGenerator(@NonNull WordEmbedding wordEmbeddings)
Instantiates a new DistributionalLexiconGeneratorDistributionalLexiconGenerator(@NonNull WordEmbedding wordEmbeddings, @NonNull Multimap<T,String> seedTerms)
Instantiates a DistributionalLexiconGeneratorDistributionalLexiconGenerator(@NonNull WordEmbedding wordEmbeddings, @NonNull Multimap<T,String> seedTerms, double threshold)
Instantiates a new DistributionalLexiconGenerator
-
Method Summary
All Methods Instance Methods Concrete Methods Modifier and Type Method Description void
addNegativeSeeds(T tag, @NonNull String... phrases)
Adds a negative example for the given tagvoid
addPositiveSeeds(T tag, @NonNull String... phrases)
Adds a positive example for the given tagMultimap<T,String>
generate()
Generate a lexicon (collection of strings) for a set of Tags
-
-
-
Constructor Detail
-
DistributionalLexiconGenerator
public DistributionalLexiconGenerator(@NonNull @NonNull WordEmbedding wordEmbeddings)
Instantiates a new DistributionalLexiconGenerator- Parameters:
wordEmbeddings
- the word embeddings
-
DistributionalLexiconGenerator
public DistributionalLexiconGenerator(@NonNull @NonNull WordEmbedding wordEmbeddings, @NonNull @NonNull Multimap<T,String> seedTerms)
Instantiates a DistributionalLexiconGenerator- Parameters:
wordEmbeddings
- the word embeddingsseedTerms
- the seed terms
-
DistributionalLexiconGenerator
public DistributionalLexiconGenerator(@NonNull @NonNull WordEmbedding wordEmbeddings, @NonNull @NonNull Multimap<T,String> seedTerms, double threshold)
Instantiates a new DistributionalLexiconGenerator- Parameters:
wordEmbeddings
- the word embeddingsseedTerms
- the seed termsthreshold
- the threshold
-
-
Method Detail
-
addNegativeSeeds
public void addNegativeSeeds(@NonNull T tag, @NonNull @NonNull String... phrases)
Adds a negative example for the given tag- Parameters:
tag
- the tagphrases
- the phrases representing a negative examples
-
addPositiveSeeds
public void addPositiveSeeds(@NonNull T tag, @NonNull @NonNull String... phrases)
Adds a positive example for the given tag- Parameters:
tag
- the tagphrases
- the phrase representing a positive example
-
generate
public Multimap<T,String> generate()
Description copied from interface:LexiconGenerator
Generate a lexicon (collection of strings) for a set of Tags- Specified by:
generate
in interfaceLexiconGenerator<T extends Tag>
- Returns:
- the multimap of tag to lemma entries
-
-