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.

Echo Liu, 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: Deborah is an Assistant Professor of Data Science at the Universita della Svizzera Italiana in Lugano. She received her Ph.D in 2023 from the Department of Statistics at the University of Oxford and was previously a postdoctoral researcher at the Barcelona School of Economics and Universitat Pompeu Fabra. Her research interests include Bayesian inference, network analysis, graphical modelling, and algorithmic and statistical fairness.

Abstract: Gaussian graphical models are widely used to infer the partial dependency structure among variables, for example, in gene studies and neuroimaging data. Under this model, conditional independence statements are encoded by zero entries in the model's precision matrix. However, the computational demands of estimating a high-dimensional precision matrix with Bayesian methods have limited the scope of applications when the number of observed variables is large. This work introduces a scalable, interpretable, and fully Bayesian method for estimating precision matrices in high-dimensional settings. Our algorithm capitalises on the sparsity of the precision matrix when the ``true'' graphical model is sparse. We also exploit the relationship between the conditional dependence structure and linear regression models to decompose the high-dimensional estimation problem via row-wise computations.  This approach enables us to parallelise some computations of posterior conditional distributions and fosters an efficient exploration of the network structure. Lastly, we characterise the efficiency of our algorithm in terms of mixing time, using newly developed concepts of conductance for the analysis of such Monte-Carlo Markov Chains algorithm.

Bio: Mariah Boudreau (she/her) is a Postdoctoral Researcher at the Roux Institute and the Network Science Institute at Northeastern University. She is a member of the MOBS Lab, and her current research focuses on mathematical and computational methods for stochastic dynamics of infectious diseases. She received a bachelor's of science from Saint Michael's College, where she studied Mathematics, and her Ph.D. from the University of Vermont, where she studied Mathematical Sciences.

Abstract: When modeling the spread of disease, heterogeneities in contact structure and stochastic modes of transmission are essential to accurately predict the size of an outbreak. Agent-based models are commonly used to capture this level of granularity, however, simulating them can be computationally expensive. Probability generating functions (PGFs) offer an efficient framework for describing stochastic transmission via branching processes. In this talk, I will introduce the PGF framework for disease spread and illustrate its usefulness through two examples. First, I will demonstrate the use of PGFs over contact networks to determine final outbreak size results, along with their sensitivity to noise. Second, I will introduce how this framework can be extended with a time-dependent PGF and track particular disease progression counts over time. Two counts of interest are cumulative infection and daily infection counts. Lastly, I will discuss future work for PGF sensitivity analyses and time-dependent PGF applications.

Bio: Gülşah is a PhD candidate in Computational Communication at UCLA. Prior to joining UCLA, she received her bachelor’s and master’s degrees in industrial engineering in her home country, Turkey. Her research explores the drivers of collective intelligence, particularly the role of diversity and networks in organizational and scientific teams, using a variety of methods including agent-based modeling, human subject experiments, and natural language processing.

Abstract: Collaboration enables teams to solve problems beyond the reach of their individual members in contexts ranging from foraging insects to high-energy physics. While communication networks play a pivotal role in team success, previous research reaches contradictory conclusions on the optimal network topology. Networks that slow information transmission help maintain diversity, leading teams to explore more of the problem space and find better solutions in the long run, while networks that maximize communication efficiency allow teams to exploit known solutions, boosting overall team performance. Many previous results are based on models in which agents use their network connections to copy better-performing team members, but the solutions found by these teams tend to be worse than those discovered by a collection of the same agents working as individuals. We develop a model of team problem solving in which, in addition to copying better performing teammates, agents with diverse perspectives can also collaborate to discover otherwise inaccessible solutions. We show that the optimal network structure depends critically on the extent to which agents collaborate versus copy. Our results resolve the apparent conflict regarding the relationship between networks and problem solving by demonstrating that when agents primarily copy one another, inefficient “exploration” networks are most effective, but when agents mostly use their connections for collaboration, more efficient “exploitation” networks win out.

Bio: Emma is a PhD student in Computational Social Sciences with Vicenç Gómez at UPF. She is also part of the Computational Social Science Lab (Universität Konstanz,TU Graz, CSH Vienna) with David Garcia as a co-supervisor. She is interested in polarization in social media, online human interaction and algorithmic fairness. Her background is in Physics and AI. Usually working with networks and data analysis.

Abstract: Political conflict is an essential element of democratic systems, but can also threaten their existence if it becomes too intense. This happens particularly when most political issues become aligned along the same major fault line, splitting society into two antagonistic camps. In the 20th century, major fault lines were formed by structural conflicts, like owners vs. workers, center vs. periphery, etc. But these classical cleavages have since lost their explanatory power. Instead of theorizing new cleavages, we present the FAULTANA (FAULT-line Alignment Network Analysis) pipeline, a computational method to uncover major fault lines in data of signed online interactions. Our method makes it possible to quantify the degree of antagonism prevalent in different online debates, as well as how aligned each debate is to the major fault line. This makes it possible to identify the wedge issues driving polarization, characterized by both intense antagonism and alignment. We apply our approach to large-scale data sets of Birdwatch, a US-based Twitter fact-checking community and the discussion forums of DerStandard, an Austrian online newspaper. We find that both online communities are divided into two large groups and that their separation follows political identities and topics. In addition, for DerStandard, we pinpoint issues that reinforce societal fault lines and thus drive polarization. We also identify issues that trigger online conflict without strictly aligning with those dividing lines (e.g. COVID-19). Our methods allow us to construct a time-resolved picture of affective polarization that shows the separate contributions of cohesiveness and divisiveness to the dynamics of alignment during contentious elections and events.

Bio: Keiko Nomura is a marine social-ecological systems scientist focused on fisheries sustainability and marine resource management. Her ocean-centered research has covered topics like marine spatial planning, small-scale fisheries adaptation, and conflict and cooperation in international fisheries. She combines quantitative network and spatial analyses with qualitative governance contexts to address key questions on ocean resource use. As a Postdoctoral Researcher at CU Boulder, she is now exploring how social-ecological factors influence seafood trade in the Pacific and, in turn, how this supports food security and sustainable development in the region.

Abstract: Aquatic food trade distributes essential nutrients from fisheries across the globe. Yet these nutrients often flow toward more secure nations, even though many fisheries are supplied by lower-income, nutrient-insecure states. These trade networks are vulnerable to environmental and socioeconomic disruptions, which can reduce fisheries production or increase reliance on imports, with disproportionate impacts on already at-risk countries. In Oceania, several Pacific Island countries (PICs) may be particularly vulnerable to negative impacts from fisheries supply chain disruptions, despite their central role in supplying global tuna. The region depends heavily on food imports, faces rising unhealthy food consumption, and experiences ongoing diet-related health challenges. PICs often act as “source” countries, supplying seafood while receiving few social or health benefits in return. In this talk, I will share both current progress and future directions. Overall, this study combines consumption and nutrient data from the Aquatic Resources in Trade Database and the Aquatic Food Composition Database to trace the flow of fisheries-derived nutrients in and out of PICs. I’ve examined the roles that countries play in these trade networks (sources, exporters, or consumers), the nutritional implications for local populations, and the structure of Pacific fisheries trade. Initial results indicate that the Pacific trade network is structurally distinct from the global network: it is sparser, more disconnected, and more asymmetric. PICs generally experience net outflows of key nutrients, especially protein, vitamin B12, and fatty acids, highlighting their role as suppliers to other regions. Yet, domestic production and consumption still account for a meaningful share of nutrient retention, revealing both vulnerability and resilience. Country-level outcomes also appear to be shaped by fisheries and trade agreements. Looking ahead, I plan to deepen this analysis through network modeling and statistical approaches to examine how social-ecological factors influence nutrient trade flows, and what health impacts may emerge from these international food systems. These insights could help inform equitable development strategies for blue food systems in the Pacific and beyond.

Bio: Anastassia Vybornova (she/her) is a postdoctoral researcher at the Copenhagen Center for Social Data Science (SODAS), University of Copenhagen. She holds a PhD in Urban Data Science, and has a background in Technical Physics, Transcultural Communication, and Environmental Science. Anastassia’s current research is dedicated to the intersection of social and spatial networks, while her current leisure time is dedicated to community organizing in micromovements of solidarity.

Abstract: Highways are physical barriers that cut opportunities for social connections, but the magnitude of this effect has not been quantified. Such quantitative evidence would enable policy-makers to prioritize interventions that reconnect urban communities - an urgent need in many contemporary cities. In this talk, I will present our recent work, ‘Urban Highways Are Barriers to Social Ties’'. We relate urban highways in the 50 largest US cities with massive, geolocated online social network data to quantify the decrease in social connectivity associated with highways. We find that this barrier effect is strong in all 50 cities, and particularly prominent over shorter distances. We also confirm this effect for highways that are historically associated with racial segregation. Our research demonstrates with high granularity the long-lasting impact of decades-old infrastructure on society and provides tools for evidence-based remedies.

Bio: Izabel Pirimai Aguiar is an Omidyar Postdoctoral Fellow at the Santa Fe Institute, where she joined this fall after completing her PhD with Johan Ugander in the Institute for Computational and Mathematical Engineering at Stanford University. Izabel works on integrating social theory into the development of network analysis tools and is generally delighted by research that exists in the intersection of interesting social theories and their challenging technical counterparts. 

Abstract: When people are asked to recall their social networks, theoretical and empirical work tells us that they rely on shortcuts, or heuristics. Cognitive Social Structures (CSS) are multilayer social networks where each layer corresponds to an individual’s perception of the network. With multiple perceptions of the same network, CSSs contain rich information about how these heuristics manifest, motivating the question, Can we identify people who share the same heuristics? In this work, we propose a method for identifying cognitive structure across multiple network perceptions, analogous to how community detection aims to identify social structure in a network. To simultaneously model the joint latent social and cognitive structure, we study CSSs as three-dimensional tensors, employing low-rank nonnegative Tucker decompositions (NNTuck) to approximate the CSS— a procedure closely related to estimating a multilayer stochastic block model (SBM) from such data. We propose the resulting latent cognitive space as an operationalization of the sociological theory of social cognition by identifying individuals who share relational schema. In addition to modeling cognitively independent, dependent, and redundant networks, we propose a specific model instance and related statistical test for testing when there is social-cognitive agreement in a network: when the social and cognitive structures are equivalent. We use our approach to analyze four different CSSs and give insights into the latent cognitive structures of those networks.

Previous talks
(Available on Youtube)