Popular pornographic websites are deemed to reinforce a male, white and heterosexual point of view and thus contribute to fostering hegemonic masculinity (Burke 2016), the sexualization of minorities (Fritz et al. 2020), and #heteronormative porn culture (Saunders 2020).
However, how exactly this happens has remained largely overlooked. At a platform level, the role of algorithms is pivotal in these processes. Algorithms contribute to managing content visibility (Bucher 2012) and can reiterate the gender #bias coded into them (Noble 2018), reifying a specific view of the world due to their social embeddedness (Pasquale 2016).
This talk analyses data collected from one of the biggest porn #streaming platforms, PornHub, to inquire how its #recommendation #system might vary according to specific socio-demographic characteristics.
To do that, we use a set of user accounts; we recreate the browsing activity of users of the website based on predetermined viewing patterns across two dimensions: gender identity and sexual interests, assigned by the researchers (Sandvig 2014). We then collected evidence on profiling taking advantage of the browser extension Pornhub Tracking Exposed.
We consider how the recommending system changes according to users’ self-declared #gender and sexual interests: this includes changes in homepage layout, recommended videos, suggested categories, and popular content.
Based on this analysis, we argue that the combination of platform affordances and algorithmic suggestions on Pornhub significantly contributes to reiterating a heteronormative perspective on sexual #desire and #sexuality typical of heterosexual, white, and hegemonic masculinity.