Facebook Dating – Network Analysis
https://www.theverge.com/2019/9/5/20850307/facebook-dating-united-states-launch
In September 2019, Facebook launched a new feature called Facebook Dating. Once Facebook users opt in to this feature, the platform suggests potential matches based on their interests and Facebook activity. Users can send a message to any profile that is suggested to them and hope for a response. One feature of particular interest is the “Secret Crush” feature. This feature allows users to list up to 9 of their Facebook friends as a “Secret Crush”, and two people will be notified if they have each other listed. Facebook emphasized that they did not want to include any game-like aspect to the feature — specifically swiping.
From a Networks point of view, it is interesting to consider how the integrated matching algorithm will use one’s friendship network to generate suggested matches. It stands to reason that the more common friendship links you share with someone the more likely you are to be matched. This application may even raise the rate of response from prospective matches compared to platforms like Tinder and Bumble, considering there is a number of calculated links tying individuals to each other. If we consider the candidates that are seeking the same gender as the R nodes and the gender being sought as the L nodes, we can evaluate this scenario using the bipartite graph.
Assuming edges only go from R to L:
S = subset of R nodes; candidates seeking the same gender of partner
N(S) = all nodes connected to S; Partners being sought after
S is a constricted set if N(S) has fewer nodes than S. S is a constricted set if there are not enough partners to match the number of candidates.
Facebook’s integrated matching algorithm should dramatically decrease the prevalence of constricted sets where there is a massive pool of candidates vying for the attention of the most “desirable” matches. It will be interesting to observe how much dating success this feature generates compared to its competitors.