The Uli Dataset: An Exercise in Experience Led Annotation of oGBV
This addresses the problem of automated detection of gendered abuse for users in the Global majority, though it is incremental as it focuses on dataset creation rather than novel detection methods.
The authors tackled the lack of language-specific and contextual data for detecting online gender-based violence by creating the Uli Dataset, which includes annotated tweets in Hindi, Tamil, and Indian English, demonstrating a participatory annotation approach involving experts from women and LGBTQIA communities in South Asia.
Online gender based violence has grown concomitantly with adoption of the internet and social media. Its effects are worse in the Global majority where many users use social media in languages other than English. The scale and volume of conversations on the internet has necessitated the need for automated detection of hate speech, and more specifically gendered abuse. There is, however, a lack of language specific and contextual data to build such automated tools. In this paper we present a dataset on gendered abuse in three languages- Hindi, Tamil and Indian English. The dataset comprises of tweets annotated along three questions pertaining to the experience of gender abuse, by experts who identify as women or a member of the LGBTQIA community in South Asia. Through this dataset we demonstrate a participatory approach to creating datasets that drive AI systems.