Girls’s sexualization on social networks: the need for semantic intelligence on Instagram

Abstract

Instagram is the millennials’ favourite social network. Its community regulations prohibit nudes, except in cases of mastectomies, breastfeeding, paintings and sculptures. Its artificial intelligence only detects percentages of skin in photographs. Lacking semantic intelligence, its algorithms do not discover other forms of hypersexualization. In this subterfuge, children and girls show their bodies dressed and reified, in school uniforms, surrounded by children's elements. They recreate the Lolita myth and get millions of followers. This paper describes this dark phenomenon to investigate: the hashtags that use this new form of pornography not censurable by the algorithms; the objective and subjective perceptions of those photographs; and a proposal for gender intervention. The reserach contains two phases: a longitudinal bibliographic exploration, which delves into Instagram and its conditions of use, personal identity and narrative, and semantic intelligence to read photographs; and an inquiry that correlates the reciprocity of hashtags, its normal denotation and its hypersexualized connotation. The results are arranged around two modalities: children who are not of legal age to have an account and girls who are of legal age to have an account. The quantitative part includes: hashtag, account name, country and followers. The qualitative part explores: description of the elements of the photographs, normal meanings and implicit pornographic meanings. It is concluded that it is necessary to: detect other hashtags that describe other pornographies; expand this prospect internationally; extrapolate this research to other social networks, such as YouTube and TikTok; share the results with social networks and the Security Forces; monitor the use of social networks by parents; training in educational settings; acquisition of semantic intelligence by social networks and minors.