This study addressed toxicity in tweets targeting Italian politicians during the propaganda campaign preceding Italian snap elections to surface differences in gender. A large sample of replies (163, 544 replies) from the 20 most popular tweets posted by 123 Italian women and 121 men politicians is considered; through network analysis, semantic analysis and natural language processing, trends in the way toxicity is crafted and distributed across communities are identified. Main takeaways are the following: women politicians are targeted more than man politicians, especially through threats or attacks to their identity; women politicians are target of attacks from the same groups of users more than man, suggesting the presence of a misogynistic group of haters; women haters like men haters tend to attack women more than man; “being a woman” is topicalized in toxic speech against women and associated to fearmongering words.

Read the FULL PAPER HERE

“We thank Reset for sharing with us insights and criticism, while discussing and debating the focus of this research. We also thank Wateronmars for their help in the process of data collection”.

Di Elena Musi, Lorenzo Federico, Ayoub Mounim e Gianni Riotta