Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
The semantic organization of the English odor vocabulary
Stockholm University, Faculty of Social Sciences, Department of Psychology, Perception and psychophysics. Stockholm University, Faculty of Humanities, Department of Linguistics, General Linguistics.ORCID iD: 0000-0003-0897-8911
Stockholm University, Faculty of Social Sciences, Department of Psychology, Perception and psychophysics.ORCID iD: 0000-0003-3418-0700
Stockholm University, Faculty of Social Sciences, Department of Psychology, Perception and psychophysics.ORCID iD: 0000-0002-0856-0569
2019 (English)Conference paper, Poster (with or without abstract) (Refereed)
Abstract [en]

Most people find it difficult to name familiar odors. Many languages, including English, lack a vocabulary devoted to describing odor qualities (compared to, e.g., a color term vocabulary), and little is known about the vocabulary used to describe odors. Attempts to establish “primary odor descriptors” have been unsuccessful. To date, research on odor vocabulary has rarely been done from a data-driven, empirical perspective.

We present a study on the semantic organization of odor vocabulary, based on the distribution of words in olfactory and gustatory contexts, using a three-billion-word corpus of written English. Using a data-driven, computational linguistic approach developed in our lab, we quantify terms with respect to the degree of olfactory-semantic content they convey. We then derive the semantic organization of the top 200 olfactory-related terms, using a distributional-semantic word vector model, which represent semantic distances as multidimensional vector distances. The model is trained on olfactory and gustatory contexts, using the word2vec neural network implementation. Based on the semantic distances, we then use dimensionality reduction and clustering techniques (i.e., PCA and hierarchical clustering) to derive a 3-dimensional, corpus-based semantic space, and six principal descriptor clusters.

Using distances based on the Draveneiks odor-term ratings data set, we also derive a semantic space with six specific clusters for the Draveneiks terms. The organization and clustering of our corpus-based semantic space match with the ratings-based semantic space, thereby showing the viability of our corpus-based approach. Based on our corpus-based data, we finally propose a novel domain-general odor term taxonomy (i.e., a domain-general odor wheel) that captures the dimensions and clusters identified in our analyses.

Place, publisher, year, edition, pages
2019.
Keywords [en]
odor semantics, odor taxonomy, corpus-based modeling, word embedding, distributional-semantic modeling, Olfactory Association Index, hierarchical clustering, PCA
National Category
General Language Studies and Linguistics Psychology (excluding Applied Psychology)
Research subject
Linguistics; Psychology
Identifiers
URN: urn:nbn:se:su:diva-172936OAI: oai:DiVA.org:su-172936DiVA, id: diva2:1351341
Conference
ECRO (European Chemoreception Research Organization), Trieste, Italy, 11-14 September, 2019
Funder
Riksbankens Jubileumsfond, M14-0375:1Available from: 2019-09-14 Created: 2019-09-14 Last updated: 2019-09-19Bibliographically approved

Open Access in DiVA

fulltext(4577 kB)5 downloads
File information
File name FULLTEXT01.pdfFile size 4577 kBChecksum SHA-512
3b77327b4144897c3143686be518861364c97c3ae22c25b0b41f36e108d15f329fdc06465a015c5ae85887c1c289512003bef4e86a04a2fcf02c9132182d4b7f
Type fulltextMimetype application/pdf

Search in DiVA

By author/editor
Hörberg, ThomasLarsson, MariaOlofsson, Jonas K.
By organisation
Perception and psychophysicsGeneral Linguistics
General Language Studies and LinguisticsPsychology (excluding Applied Psychology)

Search outside of DiVA

GoogleGoogle Scholar
Total: 5 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 61 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf