Beyond Activity Space: Detecting Communities in Ecological Networks

03/18/2019 ∙ by Wenna Xi, et al. ∙ 0

Emerging research suggests that the extent to which activity spaces -- the collection of an individual's routine activity locations -- overlap provides important information about the functioning of a city and its neighborhoods. To study patterns of overlapping activity spaces, we draw on the notion of an ecological network, a type of two-mode network with the two modes being individuals and the geographic locations where individuals perform routine activities. We describe a method for detecting "ecological communities" within these networks based on shared activity locations among individuals. Specifically, we identify latent activity pattern profiles, which, for each community, summarize its members' probability distribution of going to each location, and community assignment vectors, which, for each individual, summarize his/her probability distribution of belonging to each community. Using data from the Adolescent Health and Development in Context (AHDC) Study, we employ latent Dirichlet allocation (LDA) to identify activity pattern profiles and communities. We then explore differences across neighborhoods in the strength, and within-neighborhood consistency of community assignment. We hypothesize that these aspects of the neighborhood structure of ecological community membership capture meaningful dimensions of neighborhood functioning likely to co-vary with economic and racial composition. We discuss the implications of a focus on ecological communities for the conduct of "neighborhood effects" research more broadly.



There are no comments yet.


This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.