Detecting Sexist MEME On The Web: A Study on Textual and Visual Cues

Abstract

In recent years, it is evident the interest in the role of women within society and, in particular, the way we approach and refer to them. However, sexism as a form of discrimination towards women spread exponentially through the web and at a very high frequency, especially in the form of memes. Memes, which are typically composed of pictorial and textual components, can convey messages ranging from women stereotype, shaming, objectification to violence. In order to counterattack this phenomenon, in this paper we give a first insight in the field of automatic detection of sexist memes, by investigating both unimodal and multimodal approaches to understand the contribution of textual and visual cues.