Hate Speech on the Swedish Flashback Platform

The final speakers in this AoIR 2018 session are Emma von Essen and Joakim Jansson, whose focus is on online hate speech towards women and foreigners, and the role of anonymity in enabling the expression of such hate; her project’s interest is especially in the Swedish context, and it hopes to predict the expression of hateful ideas.

The focus of the project is on relevant fora on the Swedish platform Flashback, which emphasised the anonymity of its participants but was subsequently shown to be less secure than claimed; the project scraped posts from these fora, and coded these posts for hateful content and topical focus. From this, is it possible to predict the occurrence of hateful content?

Some one third of the users registered before March 2007 have been identified by ‘troll-hunting’ journalists, and it is therefore possible to assess the differences in the activities of these early and subsequent late adopters; as it turns out, while 7% of members of the platform post hateful content, later adopters are less likely to express hateful content (or to participate in relevant fora at all). At the same time, there is a shift in the focus of such hate; earlier adopters are more likely to express hate against foreigners, and later adopters are more likely to target women and feminists.

In particular, decreased anonymity appears to decrease the expression of hateful content, and participation in relevant fora overall. But there is also a shift in the targets of hate here.