Uncovering the Role of Weak Ties in Implicit Networks of Influence: A Network Analysis on Recommendation Algorithms

In the contemporary postmodern context, consumers are often portrayed as liberated from social ties, fostering an environment conducive to individualism. Algorithmic artifacts, such as Recommendation Algorithms (RAs), are contributing to this paradigm by functioning as anti-link tools: they establish implicit social links among individuals with similar preferences, giving rise to clusters termed neighborhoods. These neighborhoods facilitate the provision of personalized suggestions based on shared interests, paradoxically fostering social connections amid the backdrop of individualism. RAs actively generate implicit networks of influence characterized by users sharing analogous preferences, thereby enhancing the predictability of user behaviors. Despite extensive research on explicit networks of influence and the impact of RAs on decision-making, there remains a scarcity of evidence on (1) how users influence others within these implicitly generated networks, and (2) the roles they play in shaping the flow of information across such implicit networks.