Stereotypical gender actions can be extracted from Web text
This work addresses the challenge of augmenting commonsense knowledge repositories with gender stereotypes from natural text, which is incremental as it builds on existing methods for bias extraction.
The study tackled the problem of extracting gender-specific actions from text corpora and Twitter to compare them with stereotypical expectations, achieving a Spearman correlation of 0.47 with human ratings and an AUC of 0.76 for polarity prediction.
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.