What do language representations really represent?

Research output: Contribution to journalJournal articlepeer-review

Standard

What do language representations really represent? / Bjerva, Johannes; Östling, Robert; Veiga, Maria Han; Tiedemann, Jörg; Augenstein, Isabelle.

In: Computational Linguistics, Vol. 45, No. 2, 2019, p. 381-389.

Research output: Contribution to journalJournal articlepeer-review

Harvard

Bjerva, J, Östling, R, Veiga, MH, Tiedemann, J & Augenstein, I 2019, 'What do language representations really represent?', Computational Linguistics, vol. 45, no. 2, pp. 381-389. https://doi.org/10.1162/COLIa00351

APA

Bjerva, J., Östling, R., Veiga, M. H., Tiedemann, J., & Augenstein, I. (2019). What do language representations really represent? Computational Linguistics, 45(2), 381-389. https://doi.org/10.1162/COLIa00351

Vancouver

Bjerva J, Östling R, Veiga MH, Tiedemann J, Augenstein I. What do language representations really represent? Computational Linguistics. 2019;45(2):381-389. https://doi.org/10.1162/COLIa00351

Author

Bjerva, Johannes ; Östling, Robert ; Veiga, Maria Han ; Tiedemann, Jörg ; Augenstein, Isabelle. / What do language representations really represent?. In: Computational Linguistics. 2019 ; Vol. 45, No. 2. pp. 381-389.

Bibtex

@article{44b530d1f7cc473e8d3211aa333170c3,
title = "What do language representations really represent?",
abstract = "A neural language model trained on a text corpus can be used to induce distributed representations of words, such that similar words end up with similar representations. If the corpus is multilingual, the same model can be used to learn distributed representations of languages, such that similar languages end up with similar representations. We show that this holds even when the multilingual corpus has been translated into English, by picking up the faint signal left by the source languages. However, just as it is a thorny problem to separate semantic from syntactic similarity in word representations, it is not obvious what type of similarity is captured by language representations. We investigate correlations and causal relationships between language representations learned from translations on one hand, and genetic, geographical, and several levels of structural similarity between languages on the other. Of these, structural similarity is found to correlate most strongly with language representation similarity, whereas genetic relationships—a convenient benchmark used for evaluation in previous work—appears to be a confounding factor. Apart from implications about translation effects, we see this more generally as a case where NLP and linguistic typology can interact and benefit one another.",
author = "Johannes Bjerva and Robert {\"O}stling and Veiga, {Maria Han} and J{\"o}rg Tiedemann and Isabelle Augenstein",
year = "2019",
doi = "10.1162/COLIa00351",
language = "English",
volume = "45",
pages = "381--389",
journal = "Computational Linguistics",
issn = "1530-9312",
publisher = "MIT Press",
number = "2",

}

RIS

TY - JOUR

T1 - What do language representations really represent?

AU - Bjerva, Johannes

AU - Östling, Robert

AU - Veiga, Maria Han

AU - Tiedemann, Jörg

AU - Augenstein, Isabelle

PY - 2019

Y1 - 2019

N2 - A neural language model trained on a text corpus can be used to induce distributed representations of words, such that similar words end up with similar representations. If the corpus is multilingual, the same model can be used to learn distributed representations of languages, such that similar languages end up with similar representations. We show that this holds even when the multilingual corpus has been translated into English, by picking up the faint signal left by the source languages. However, just as it is a thorny problem to separate semantic from syntactic similarity in word representations, it is not obvious what type of similarity is captured by language representations. We investigate correlations and causal relationships between language representations learned from translations on one hand, and genetic, geographical, and several levels of structural similarity between languages on the other. Of these, structural similarity is found to correlate most strongly with language representation similarity, whereas genetic relationships—a convenient benchmark used for evaluation in previous work—appears to be a confounding factor. Apart from implications about translation effects, we see this more generally as a case where NLP and linguistic typology can interact and benefit one another.

AB - A neural language model trained on a text corpus can be used to induce distributed representations of words, such that similar words end up with similar representations. If the corpus is multilingual, the same model can be used to learn distributed representations of languages, such that similar languages end up with similar representations. We show that this holds even when the multilingual corpus has been translated into English, by picking up the faint signal left by the source languages. However, just as it is a thorny problem to separate semantic from syntactic similarity in word representations, it is not obvious what type of similarity is captured by language representations. We investigate correlations and causal relationships between language representations learned from translations on one hand, and genetic, geographical, and several levels of structural similarity between languages on the other. Of these, structural similarity is found to correlate most strongly with language representation similarity, whereas genetic relationships—a convenient benchmark used for evaluation in previous work—appears to be a confounding factor. Apart from implications about translation effects, we see this more generally as a case where NLP and linguistic typology can interact and benefit one another.

UR - http://www.scopus.com/inward/record.url?scp=85081180389&partnerID=8YFLogxK

U2 - 10.1162/COLIa00351

DO - 10.1162/COLIa00351

M3 - Journal article

AN - SCOPUS:85081180389

VL - 45

SP - 381

EP - 389

JO - Computational Linguistics

JF - Computational Linguistics

SN - 1530-9312

IS - 2

ER -

ID: 239206885