The Women in Network Science (WiNS) seminar is an interdisciplinary seminar with the aim to promote and showcase research by women and nonbinary researchers in network science. 

The seminar is open to everyone. Please join the mailing list to receive announcements and zoom links for upcoming seminar talks.

Elena Candellone, Haily Merritt and Gülşah Akçakır convene this seminar series. Please get in touch if you are interested in presenting in the seminar or if you would like to nominate speakers.

For all scheduled talks, relevant preprints are available on our ZeroDivZero repository. Recordings of past talks can also be found in the ZeroDivZero repository and on Youtube.

Scheduled talks
Mondays - 11 am Eastern Time - on Zoom 

Bio: TBD

Abstract: Comparing a network to number of random networks can reveal structures of the network that are non-trivial, i.e., not caused by randomness. In this talk, we study two algorithms to sample random hypergraphs, both suited for undirected and directed hypergraphs. The algorithms are Markov chain Monte Carlo algorithms, so we consider the underlying Markov chain and experimentally compare the mixing times of the two algorithms, which is related to the speed of the algorithms. I will briefly show an application of the algorithms to a chemical reaction network, which we model as a directed hypergraph. This talk is based on the article https://doi.org/10.1093/comnet/cnaf007, which is joined work with Clara Stegehuis.

Bio: Sanjukta's research interests are multidisciplinary, with the goal of developing computational and mathematical tools to answer questions about real world physical, social, and biological systems. She received her PhD from the University of Maryland, College Park. She was a UC Presidential postdoc with a joint appointment at UC Berkeley and UCLA. She enjoys traveling and has lived on four continents. In her free time she enjoys dance, art, hiking, and attempting to climb mountains.

Abstract: TBD

Bio: TBD

Abstract: Network science offers analytical frameworks to model the brain as a system of interconnected regions, where segregated modules can be identified by optimizing the weights of pairwise connections within them. However, focusing solely on pairwise connections may be insufficient, as brain function also arises from higher-order interactions involving multiple brain areas simultaneously. Here, we leverage information theory to propose an algorithm for community detection that accounts for multivariate interactions. Our algorithm uses Total Correlation (TC) as a measure of redundancy and, starting from a random partition, iteratively reassigns nodes to maximize within-module TC, yielding optimally redundant modules. Applied to the Human Connectome Project data, the algorithm revealed modules that only partially overlap with those identified by conventional methods, and uncovered a new organization of the transmodal cortex. We also found a spatial scale where within-module redundancy and between-module synergy are optimally balanced, thus capturing a higher-order representation of the interplay between segregation and integration. Furthermore, we mapped brain regions with high and low topological specialization based on their contributions to within- or between-module redundancy. Finally, we observed age-related reconfigurations of this redundant modular structure. These results suggest unexplored topological properties of brain functional networks that can be only revealed by multivariate measures of information. Further analysis will elucidate their role in cognition and behavior.

Bio: TBD

Abstract: TBD

Bio: TBD

Abstract: TBD

Bio: TBD

Abstract: The structure of networks on social media platforms influences how information spreads, how attention is allocated, and how users might influence each other. However, understanding how these networks take shape over time, whether at a micro scale or from a more global perspective, requires unique forms of (meta)data that are often difficult to produce and parse. In this talk, I provide two ways of looking into the forces that shape networks: in one project, I provide an empirical example of the impact a single influential user’s amplification can have on following dynamics and, in the other, I present a dataset with extensive potential for illuminating the effects of higher-order endorsement patterns on network evolution. First, I explain how amplification and triad transitivity in directed social media networks are linked via a type of user called an attention broker – an influential user whose amplification choices guide the formation of attention patterns in systems where amplification with attribution occurs. I provide empirical proof for a causal relationship between amplification and follower accumulation on Twitter/X: when an account is retweeted by an attention broker, the attention broker’s audience follows the retweeted account at a higher rate because the attention broker amplified the original poster. Second, I introduce a novel dataset that bridges higher-order and dyadic networks on Bluesky. My collaborators and I collected over 300k starter packs and 1.7 billion timestamped following ties between users, ultimately producing two anonymized datasets that are currently available at the Social Media Archive at ICPSR. I present preliminary analyses of these datasets, explaining how they are reflective of platform affordances and relevant off-platform events. By presenting these projects together, I hope to generate discussion and, perhaps, future work on the ways that endorsement affordances influence the ways that networks take shape on social media platforms.

Bio: TBD

Abstract: TBD

Bio: TBD

Abstract: TBD

Previous talks
(Available on Youtube)