org.apache.ctakes.ytex.kernel.wsd

Interface WordSenseDisambiguator

All Known Implementing Classes:

WordSenseDisambiguatorImpl

public interface WordSenseDisambiguator

Method Summary

Modifier and Type	Method and Description
String	disambiguate(List<Set<String>> sentenceConcepts, int index, Set<String> contextConcepts, int windowSize, ConceptSimilarityService.SimilarityMetricEnum metric, Map<String,Double> scoreMap)
String	disambiguate(List<Set<String>> sentenceConcepts, int index, Set<String> contextConcepts, int windowSize, ConceptSimilarityService.SimilarityMetricEnum metric, Map<String,Double> scoreMap, boolean weighted) Disambiguate a named entity.

Method Detail

disambiguate

String disambiguate(List<Set<String>> sentenceConcepts,
                    int index,
                    Set<String> contextConcepts,
                    int windowSize,
                    ConceptSimilarityService.SimilarityMetricEnum metric,
                    Map<String,Double> scoreMap)

disambiguate

String disambiguate(List<Set<String>> sentenceConcepts,
                    int index,
                    Set<String> contextConcepts,
                    int windowSize,
                    ConceptSimilarityService.SimilarityMetricEnum metric,
                    Map<String,Double> scoreMap,
                    boolean weighted)

Disambiguate a named entity.

Parameters:

sentenceConcepts - named entities from the document, represented as list of sets of concept ids

index - index of target named entity to disambiguate

contextConcepts - context concepts, e.g. from title

windowSize - number of named entities on either side of target to use for disambiguation

metric - metric to use

scoreMap - optional to get the scores assigned to each concept

weighted - to weight context concepts by frequency

Returns:

highest scoring concept, or null if none of the target concepts are in the concept graph, or if all the target concepts have the same score