Publication

Bleaching Text: Abstract Features for Cross-lingual Gender Prediction

van der Goot, R., Ljubešic, N., Matroos, I., Nissim, M. & Plank, B., Jul-2018. 7 p.

Research output: Contribution to conferencePaperAcademic

APA

van der Goot, R., Ljubešic, N., Matroos, I., Nissim, M., & Plank, B. (2018). Bleaching Text: Abstract Features for Cross-lingual Gender Prediction. Paper presented at 56th Annual Meeting of the Association for Computational Linguistics, Melbourne, Australia.

Author

van der Goot, Rob ; Ljubešic, Nikola ; Matroos, Ian ; Nissim, Malvina ; Plank, Barbara. / Bleaching Text : Abstract Features for Cross-lingual Gender Prediction. Paper presented at 56th Annual Meeting of the Association for Computational Linguistics, Melbourne, Australia.7 p.

Harvard

van der Goot, R, Ljubešic, N, Matroos, I, Nissim, M & Plank, B 2018, 'Bleaching Text: Abstract Features for Cross-lingual Gender Prediction', Paper presented at 56th Annual Meeting of the Association for Computational Linguistics, Melbourne, Australia, 16/07/2018 - 20/07/2018.

Standard

Bleaching Text : Abstract Features for Cross-lingual Gender Prediction. / van der Goot, Rob; Ljubešic, Nikola; Matroos, Ian; Nissim, Malvina; Plank, Barbara.

2018. Paper presented at 56th Annual Meeting of the Association for Computational Linguistics, Melbourne, Australia.

Research output: Contribution to conferencePaperAcademic

Vancouver

van der Goot R, Ljubešic N, Matroos I, Nissim M, Plank B. Bleaching Text: Abstract Features for Cross-lingual Gender Prediction. 2018. Paper presented at 56th Annual Meeting of the Association for Computational Linguistics, Melbourne, Australia.


BibTeX

@conference{047974e36e1441edbb46ad2a79aec4a1,
title = "Bleaching Text: Abstract Features for Cross-lingual Gender Prediction",
abstract = "Gender prediction has typically focused on lexical and social network features, yielding good performance, but making systems highly language-, topic-, and platform dependent. Cross-lingual embeddings circumvent some of these limitations, but capture gender-specific style less. We propose an alternative: bleaching text, i.e., transforming lexical strings into more abstract features. This study provides evidence that such features allow for better transfer across languages. Moreover, we present a first study on the ability of humans to perform cross-lingual gender prediction. We find that human predictive power provessimilar to that of our bleached models, and both perform better than lexical models.",
author = "{van der Goot}, Rob and Nikola Ljubešic and Ian Matroos and Malvina Nissim and Barbara Plank",
year = "2018",
month = "7",
language = "English",
note = "56th Annual Meeting of the Association for Computational Linguistics, ACL 2018 ; Conference date: 16-07-2018 Through 20-07-2018",
url = "https://acl2018.org/",

}

RIS

TY - CONF

T1 - Bleaching Text

T2 - Abstract Features for Cross-lingual Gender Prediction

AU - van der Goot, Rob

AU - Ljubešic, Nikola

AU - Matroos, Ian

AU - Nissim, Malvina

AU - Plank, Barbara

PY - 2018/7

Y1 - 2018/7

N2 - Gender prediction has typically focused on lexical and social network features, yielding good performance, but making systems highly language-, topic-, and platform dependent. Cross-lingual embeddings circumvent some of these limitations, but capture gender-specific style less. We propose an alternative: bleaching text, i.e., transforming lexical strings into more abstract features. This study provides evidence that such features allow for better transfer across languages. Moreover, we present a first study on the ability of humans to perform cross-lingual gender prediction. We find that human predictive power provessimilar to that of our bleached models, and both perform better than lexical models.

AB - Gender prediction has typically focused on lexical and social network features, yielding good performance, but making systems highly language-, topic-, and platform dependent. Cross-lingual embeddings circumvent some of these limitations, but capture gender-specific style less. We propose an alternative: bleaching text, i.e., transforming lexical strings into more abstract features. This study provides evidence that such features allow for better transfer across languages. Moreover, we present a first study on the ability of humans to perform cross-lingual gender prediction. We find that human predictive power provessimilar to that of our bleached models, and both perform better than lexical models.

M3 - Paper

ER -

ID: 61170687