Are Distributional Representations Ready for the Real World? Evaluating Word Vectors for Grounded Perceptual Meaning

Li Lucy, Jon Gauthier


Abstract
Distributional word representation methods exploit word co-occurrences to build compact vector encodings of words. While these representations enjoy widespread use in modern natural language processing, it is unclear whether they accurately encode all necessary facets of conceptual meaning. In this paper, we evaluate how well these representations can predict perceptual and conceptual features of concrete concepts, drawing on two semantic norm datasets sourced from human participants. We find that several standard word representations fail to encode many salient perceptual features of concepts, and show that these deficits correlate with word-word similarity prediction errors. Our analyses provide motivation for grounded and embodied language learning approaches, which may help to remedy these deficits.
Anthology ID:
W17-2810
Volume:
Proceedings of the First Workshop on Language Grounding for Robotics
Month:
August
Year:
2017
Address:
Vancouver, Canada
Venues:
RoboNLP | WS
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
76–85
Language:
URL:
https://aclanthology.org/W17-2810
DOI:
10.18653/v1/W17-2810
Bibkey:
Cite (ACL):
Li Lucy and Jon Gauthier. 2017. Are Distributional Representations Ready for the Real World? Evaluating Word Vectors for Grounded Perceptual Meaning. In Proceedings of the First Workshop on Language Grounding for Robotics, pages 76–85, Vancouver, Canada. Association for Computational Linguistics.
Cite (Informal):
Are Distributional Representations Ready for the Real World? Evaluating Word Vectors for Grounded Perceptual Meaning (Lucy & Gauthier, 2017)
Copy Citation:
PDF:
https://aclanthology.org/W17-2810.pdf
Code
 lucy3/Graphs-Embeddings