Introduction

Terms in a natural language may be ambiguous, i.e. can be mapped to multiple distinct concepts. For example, the word ‘cold’ can refer to the viral infection ‘common cold’ or the ‘sensation of cold’. YTEX implements the 'adapted lesk' method that uses semantic similarity measures to quantify how well a concept ‘fits’ in a given context. This page describes the WSD algorithm, the configuration for the SenseDisambiguatorAnnotator, and describes how to reproduce the results of our evaluation on the NLM WSD and MSH WSD data sets.

...

Note that you must perform the additional YTEX installation tasks to use this component.

SenseDisambiguatorAnnotator

The SenseDisambiguatorAnnotator is an UIMA annotator integrated with cTAKES. cTAKES identifes named entities (EntityMention Annotations), which in turn can contain multiple concepts (OntologyConcept Feature Structures). The SenseDisambiguatorAnnotator disambiguates each ambiguous term (i.e. EntityMention with multiple OntologyConcepts) in a document as follows:

...