Towards Trauma-Informed Data Donation of Sexual Experience in Online Dating to Improve Sexual Risk Detection AI

Wenqi Zheng, Emma Walquist, Isha Datey, Xiangyu Zhou, Kelly Berishaj, Melissa Mcdonald, Michele Parkhill, Dongxiao Zhu, Douglas Zytko

Abstract: Sexual risk detection AI has been touted as a scalable solution for computer-mediated sexual violence. Data donation is a user-centered approach to producing ecologically valid datasets for sexual risk detection AI: voluntarily providing personal data that is representative of risk. However, the act of donating intimate sexual experience data can itself be traumatizing. We propose Ube: a trauma-informed sexual experience data donation app for online daters that is developed jointly with sexual violence experts and care practitioners. Cognitive walkthroughs of Ube with these experts elucidated several design approaches to mitigate retraumatization during data donation, including a conversational agent and mental health checks.