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 proves similar to that of our bleached models, and both perform better than lexical models.
Cite
@article{arxiv.1805.03122,
title = {Bleaching Text: Abstract Features for Cross-lingual Gender Prediction},
author = {Rob van der Goot and Nikola Ljubešić and Ian Matroos and Malvina Nissim and Barbara Plank},
journal= {arXiv preprint arXiv:1805.03122},
year = {2018}
}
Comments
Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics