Date: Wednesday, March 29th, 2023
9:00 am – 10:00 am Pacific Time
12:00 pm – 1:00 pm Eastern Time
Location: Weekly Seminar, Zoom
Title: Breaking Inequality Barriers Through Mutual Connections: The Desegregating Power of Triadic Closure
Abstract:
Social structure underlies access to information, resources, and opportunities and therefore can exacerbate inequality across a population in the case of network segregation. As individuals tend to form social ties with others who are similar to themselves, segregation can naturally arise. This phenomenon known as homophily is a pervasive and organic force. Designing effective interventions to push back against homophily requires identifying other counterbalancing social processes. Triadic closure, a mechanism in which two individuals with a mutual connection form a tie which each other, can be one such process. In this talk, I will challenge a long-held belief that triadic closure works in tandem with homophily. I will demonstrate the desegregating power of triadic closure through a theoretical study of fundamental network models supported by empirical evidence. I will conclude with a discussion on implications for the design of interventions in settings where individuals arrive in an online fashion, and the designer can influence the initial set of connections.
Based on a joint work with Rediet Abebe, Nicole Immorlica, Jon Kleinberg, and Brendan Lucier.
Bio:
Ali Shirali is a Ph.D. student at UC Berkeley Computer Science, advised by Rediet Abebe and Moritz Hardt. Ali’s research contributes to the scientific foundation of machine learning from a social perspective. His work has covered how humans are involved in the life-cycle of machine learning, from data annotation to evaluation, and how humans are affected by algorithmic systems, specifically, the interplay of algorithms and inequality.