Attraction Behind Beauty:

Revealing Gay Men’s Self-Presentation on a Dating App with Computer Vision

Abstract

Photos, a fundamental construct of online self-presentation, are crucial to the success of building online relationships. This paper applies social bots techniques to collect user profile photos on a Chinese gay dating app Finka and uses computer vision algorithms to analyze the selectively presented social cues from the photos. Extending from existing research on physical (i.e., facial and body cues) visual cues, we incorporate non-physical (i.e., background scenes) cues as another theoretical dimension. Results show that although showing the face and making facial expressions are not associated with user popularity, appealing faces and body disclosure are attractiveness determinators, confirming the existence of “the gay male beauty myth” in the online dating community. Interestingly, we also discovered an “ambiance effect” in profile photos: selectively choosing background scenes can act as a new self-presentation tactic that could influence a user’s popularity. Particularly, intimate scenes (e.g., home, hotel, and gym), as the most attractive non-physical cues, can arouse the desire for further interaction with the presenter. Our findings suggest that intimate background scenes can reinforce the presenters’ physical attractiveness and invite intimate relationships on dating platforms.