Stereotypical gender actions can be extracted from Web text

TL;DR

This study demonstrates that natural language data from Web texts and Twitter can be used to identify and quantify stereotypical gender actions, enhancing commonsense knowledge bases with gender bias information.

Contribution

The paper introduces a method to extract gender-specific actions from text corpora and Twitter, and validates it against human judgments, providing datasets and demonstrating the feasibility of augmenting commonsense knowledge with gender stereotypes.

Findings

Achieved a Spearman correlation of 0.47 with human judgments.

Obtained an area under the ROC curve of 0.76 for predicting gender bias.

Created datasets of 441 human-rated and 21,442 automatically-rated actions.

Abstract

We extracted gender-specific actions from text corpora and Twitter, and compared them to stereotypical expectations of people. We used Open Mind Common Sense (OMCS), a commonsense knowledge repository, to focus on actions that are pertinent to common sense and daily life of humans. We use the gender information of Twitter users and Web-corpus-based pronoun/name gender heuristics to compute the gender bias of the actions. With high recall, we obtained a Spearman correlation of 0.47 between corpus-based predictions and a human gold standard, and an area under the ROC curve of 0.76 when predicting the polarity of the gold standard. We conclude that it is feasible to use natural text (and a Twitter-derived corpus in particular) in order to augment commonsense repositories with the stereotypical gender expectations of actions. We also present a dataset of 441 commonsense actions with human judges' ratings on whether the action is typically/slightly masculine/feminine (or neutral), and another larger dataset of 21,442 actions automatically rated by the methods we investigate in this study.