Understanding, Detecting, and Mitigating Online Misogyny Against Politically Active Women

Abstract

Billions of people use social media every day. Many of them discuss political topics online. Radicalization, extreme speech, and in particular online misogyny against politically active women have become alarming negative features of online discussions. In this interdisciplinary project, we will employ mixed-methods approaches to three case studies in Germany, India, and Brazil to better understand the content and dynamics of online misogyny against politically active women and to develop methods for early identification of such emerging dynamics. We will collaborate with subject matter experts in India and Brazil as well as with media partners and affected female politicians. With citizen social science tools we will involve the general public in the process of identifying emerging campaigns of online misogyny against politically active women. This project will also develop policy briefs and regulatory approaches to address online misogyny.

Goals

Identifying online hate speech in general, and misogyny, in particular, is a new significant research challenge. While some approaches already exist, they have two important shortcomings. First, the most common approaches focus on English written texts and (mostly) on US datasets. This project will develop new standards for detecting misogyny in German, Hindi , and Portuguese. Second, while there exists a global misogynistic culture online, there are enormous local differences in the way politically active women are attacked online, which are based on local cultures, prevailing social norms, and languages. Our project tries to incorporate and connect the global and the local levels of online misogyny. Our findings will contribute to a more nuanced and context-specific understanding of misogynistic narratives against female politicians, which is essential for developing automated detection systems. Another goal of this collaboration is to further develop text and topic analytical methods for studying data collections of online misogyny in an international setting.

Online polarization, fake news, and extreme speech are core challenges of our time. Researchers, regulators, and platform providers, but also politicians are pushed to aggressively tackle these challenges because finding (at least partial) solutions is essential for two reasons. First, the future of our online interactions in healthy and open online communities heavily depends on whether we can mitigate the rampant proliferation of online hate. Secondly, the internet and social media play such an important role in our lives that the future of our society and the future of a healthy democracy will depend more and more on a functioning and healthy online ecosystem. In this project, we hope to contribute towards this goal through a better understanding and faster detection of online misogyny.