Dating platforms provide the opportunity to indicate interest in other users by reacting to their online profile. In a binary selection procedure, one may either like (swipe right) or reject (swipe left) a given account. Thereby, every profile picture (regardless of whether rated attractive or not) should be associated with a specific increase in bodily arousal on the recipient’s side. Various studies have shown pupil diameter to constitute a valid indicator of changes in bodily activation. Moreover, recent research on machine learning techniques provides promising results on classifying emotional states with the help of pupillary dynamics. The current project aims to implement a pupil-based algorithm to automatically detect user preferences while viewing faces of potential partners.