Suchergebnisse

Charles J. Gomez on The Quantified Scholar and Study Gods.

Modern scientific research evokes ecological imagery and metaphors, given that it is global, interdependent, and diverse. Ecological network structures—like matrices of species inhabiting islands across an archipelago—can be reordered to form nested patterns. These patterns describe the overall health of ecosystems, place species on a spectrum between being described as generalists (foxes) or specialists (hedgehogs), and which of these interactions might appear or disappear. Using the number of citations universities receive for work published in a particular subfield taken from over 20 million scientific publications in OpenAlex, we construct and analyze yearly nested ecological networks of a dozen academic fields between 1990 to 2017. We find increasingly nested structures across fields, indicating a robust global research ecology that also infers future acknowledgment in different subfields. We argue that this framework can inform policy on scientific research and university funding and evaluation.

Collective problem-solving networks are common in modern life. They often benefit from having diverse members with complementary skills and perspectives, however the benefits of information exchange and synthesis among them may be squandered if they self-select away from diverse counterparts and towards homogeneous groups or perceived competency. Building on the extensive tradition of "exploration and exploitation" agent-based models (ABMs), we initialize communicative networks with diverse groups of agents who solve a complex problem represented by Kauffman's NK problem space. We compare three types of ABMs, where the initial network and agent setups are identical but differentiated by agents' attachment proclivities: (1) diversity- seeking networks, where agents prefer ties with different-agents; (2) homophily-seeking networks, where agents prefer ties with similar-agents; and (3) merit-seeking networks, where agents prefer ties with agents who have found better solutions. We find that diversity-seeking networks perform well because diversity promotes more exploration for solutions, but it also fosters network structures that disseminate these higher quality solutions more effectively than merit-seeking and homophily-seeking networks.