From Word to Sense Embeddings: A Survey on Vector Representations of Meaning

  • 2018-05-10 15:56:48
  • Jose Camacho-Collados, Taher Pilehvar
  • 36

Abstract

Over the past years, distributed representations have proven effective andflexible keepers of prior knowledge to be integrated into downstreamapplications. This survey is focused on semantic representation of meaning. Westart from the theoretical background behind word vector space models andhighlight one of its main limitations: the meaning conflation deficiency. Then,we explain how this deficiency can be addressed through a transition from wordlevel to the more fine-grained level of word senses (in its broaderacceptation) as a method for modelling unambiguous lexical meaning. We presenta comprehensive overview of the wide range of techniques in the two mainbranches of sense representation, i.e., unsupervised and knowledge-based.Finally, this survey covers the main evaluation procedures and an analysis offive important aspects: interpretability, sense granularity, adaptability todifferent domains, compositionality and integration into downstreamapplications.

 

Introduction (beta)

None

 

Conclusion (beta)

None