Work

What Company Do Semantically Ambiguous Words Keep? Insights from Distributional Word Vectors.

Year: 2018

Type: conference-paper

Abstract: The diversity of a word's contexts affects its acquisition and processing.Can differences between word types such as monosemes (unambiguous words), polysemes (multiple related senses), and homonyms (multiple unrelated meanings) be related to distributional properties of these words?We tested for traces of number and relatedness of meaning in vector representations by comparing the distance between words of each type and vector representations of various "contexts": their dictionary defnitions (an extreme disambiguating context), their use in flm subtitles (a natural context), and their semantic neighbours in vector space (a vector-space-internal context).Whereas dictionary defnitions reveal a three-way split between our word types, the other two contexts produced a two-way split between ambiguous and unambiguous words.These inconsistencies align with some discrepancies in behavioural studies and present a paradox regarding how models learn meaning relatedness despite natural contexts seemingly lacking such relatedness.We argue that viewing ambiguity as a continuum could resolve many of these issues.

Source: TSpace (University of Toronto)

Authors: Barend Beekhuizen, Saša Milić, Blair C. Armstrong, Suzanne Stevenson

Language: English

Topic: Natural Language Processing Techniques

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

Funder: Natural Sciences and Engineering Research Council of Canada

Awards: G279537525, Probabilistic Models of Semantic and Pragmatic Acquisition and Processing, Advancing Neural Network Models of Language Processing

OpenAlex ID: W2940617875

FWCI: 0.633

Cites: 24

Cited by: 8

Related to: 10