Semantic projection recovers rich human knowledge of multiple object features from word embeddings
Name
Semantic projection recovers rich human knowledge of multiple object features from word embeddings.pdf
Description
Accepted version
Size
2.56 MB
Format
Adobe PDF
Checksum (MD5)
0161392ec03174ab5086532a8512fe4f
Author(s)
Grand, Gabriel
Blank, Idan Asher
Pereira, Francisco
Fedorenko, Evelina
Date Issued
2022
Journal
Nature Human Behaviour
Publisher
Springer Science and Business Media LLC
Citation
Grand, Gabriel, Blank, Idan Asher, Pereira, Francisco and Fedorenko, Evelina. 2022. "Semantic projection recovers rich human knowledge of multiple object features from word embeddings." Nature Human Behaviour, 6 (7).
Version
Author's final manuscript
Abstract
How is knowledge about word meaning represented in the mental lexicon? Current computational models infer word meanings from lexical co-occurrence patterns. They learn to represent words as vectors in a multidimensional space, wherein words that are used in more similar linguistic contexts-that is, are more semantically related-are located closer together. However, whereas inter-word proximity captures only overall relatedness, human judgements are highly context dependent. For example, dolphins and alligators are similar in size but differ in dangerousness. Here, we use a domain-general method to extract context-dependent relationships from word embeddings: 'semantic projection' of word-vectors onto lines that represent features such as size (the line connecting the words 'small' and 'big') or danger ('safe' to 'dangerous'), analogous to 'mental scales'. This method recovers human judgements across various object categories and properties. Thus, the geometry of word embeddings explicitly represents a wealth of context-dependent world knowledge.
MIT Department
Massachusetts Institute of Technology. Department of Brain and Cognitive Sciences
Terms of Use
Creative Commons Attribution-Noncommercial-Share Alike
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1038/S41562-022-01316-8
