Winter 2021 PROGRAM (University of Washington)

Bio: Sofia Teixeira received her PhD in Information Systems and Computer Engineering from Universidade de Lisboa (Portugal). She was a postdoc at Indiana University between 2019 and 2020, and before that she was a junior researcher at IDSS group/INESC-ID since 2010. Currently, she is a Research Scientist at Hospital da Luz Learning Health, Luz Lisboa. She is also the Chair of the advisory board of yrCSS, and a member of the Council of the CSS. Sofia's research interests include modelling and analyzing different complex systems through the development of new algorithms on graphs and the application of network science and machine learning methods. Some fields of interest are Network Neuroscience, Evolutionary Game Theory, Cognitive Science and Computational Epidemiology, among others.

Abstract: From social contracts to climate agreements, individuals engage in groups that must collectively reach decisions with varying levels of equality and fairness. These dilemmas also pervade Distributed Artificial Intelligence, in domains such as automated negotiation, conflict resolution or resource allocation. As evidenced by the well-known Ultimatum Game -- where a Proposer has to divide a resource with a Responder -- payoff-maximizing outcomes are frequently at odds with fairness. Eliciting equality in populations of self-regarding agents requires judicious interventions. Here we use knowledge about agents' social networks to implement fairness mechanisms, in the context of Multiplayer Ultimatum Games. We focus on network-based role assignment and show that preferentially attributing the role of Proposer to low-connected nodes increases the fairness levels in a population. We evaluate the effectiveness of low-degree Proposer assignment considering networks with different average connectivity, group sizes, and group voting rules when accepting proposals (e.g. majority or unanimity). We further show that low-degree Proposer assignment is efficient, not only optimizing fairness, but also the average payoff level in the population. Finally, we show that stricter voting rules (i.e., imposing an accepting consensus as a requirement for collectives to accept a proposal) attenuates the unfairness that results from situations where high-degree nodes (hubs) are the natural candidates to play as Proposers. Our results suggest new routes to use role assignment and voting mechanisms to prevent unfair behaviors from spreading on complex networks.

Website: www.andreiasofiateixeira.com

Bio: Jessica Davis is a fourth year PhD student in Network Science at Northeastern University in the MOBS Lab. She completed her undergraduate degree in applied mathematics and communication studies at the University of North Carolina at Chapel Hill. Broadly speaking, she is interested in understanding how complex socio-technical systems affect the spread of information, behavior, and infectious diseases. Her current research examines how mechanistic models can be used to provide insight into the emerging dynamics of infectious diseases.

Abstract: As an initial response to the emerging SARS-CoV-2 virus, many countries adopted a policy of only testing symptomatic individuals with a travel history linked to China. The narrowness of this criteria resulted in detecting only a small fraction of symptomatic cases, leaving many countries unaware of the local, domestic transmission. We implement a data-driven, stochastic, spatial, and age-structured epidemic model that relies on the global mobility network to provide a statistical analysis of the cryptic transmission phase and the ensuing first wave of the COVID-19 pandemic. We find that wide-spread community transmission was likely occurring in several countries in Europe and states in the US as early as January 2020. Our results indicate that international travel was the key driver of the early introduction of SARS-CoV-2 with possible importation and transmission events as early as December 2019. However, internal importations (within the US and European continent) were critical in the sustained propagation of the virus after many countries implemented travel restrictions in January and February. Analyzing the initial cryptic transmission phase provides insight into the early spreading dynamics of the virus and, in combination with sequencing data from SARS-CoV-2 genomes, can be used to reconstruct the early epidemic history of the COVID-19 pandemic in greater detail. Leveraging the methods used in this analysis can provide a statistical picture to assess the risk of emerging viruses and their variants when data are not available.

Website: https://www.networkscienceinstitute.org/people/jessica-davis

Bio: Sandra González-Bailón is an Associate Professor at the Annenberg School for Communication, and affiliated faculty at the Warren Center for Network and Data Sciences. Prior to joining Penn, she was a Research Fellow at the Oxford Internet Institute (2008-2013). She completed her doctoral degree in Nuffield College (link is external) (University of Oxford) and her undergraduate studies at the University of Barcelona. Her research lies at the intersection of network science, data mining, computational tools, and political communication. Her applied research looks at how online networks shape exposure to information, with implications for how we think about political engagement, mobilization dynamics, information diffusion, and news consumption.

Abstract: Information manipulation is widespread in today's media environment. Online networks have disrupted the gatekeeping role of traditional media by allowing various actors to influence the public agenda; they have also allowed automated accounts (or bots) to blend with human activity in the flow of information. In this talk, I will discuss evidence assessing the impact that bots have on the dissemination of content during contentious political events evolving in real-time on social media. I will focus on events of heightened political tension because they are particularly susceptible to information campaigns designed to mislead or exacerbate conflict. This research employs tools from network science, natural language processing, and machine learning to analyze the diffusion structure, the content of the messages diffused, and the actors behind those messages as the political events unfolded. I will show that verified accounts are significantly more visible than unverified bots in the coverage of the events but also that bots attract more attention than human accounts. The findings highlight that social media and the web are very different news ecosystems in terms of prevalent news sources and that both humans and bots contribute to generate discrepancy in news visibility with their activity. They also highlight the importance of clearly defining the policies that underlie the verification process of social media accounts.

Website: https://dimenet.asc.upenn.edu/

Bio: Nina Cesare is a postdoctoral associate at Boston University's School of Public Health. received her Ph.D. in sociology from the University of Washington. Her research explores the use of digital data and the spatial and social complexity of social determinants of health. She strives to create work that bridges data science and sociology.

Abstract: The digital traces we leave online offer unprecedented insight into the health-related attitudes and behaviors of a population. Each source offers unique advantages and challenges that the researcher must account for when designing a study. In this talk, I will discuss how my colleagues and I have used Twitter and Google Trends to assess trends pertaining to physical inactivity, pregnancy loss, and COVID-19 misinformation. I will address the challenges of using these and other digital sources, focusing primarily on the limitations of using semi-structured, self-reported information, and on the ethical implications of using open data to monitor collective health.

Website: http://ninacesare.com/

Bio: Ekaterina (Kath) Landgren is an Applied Mathematics Ph.D. candidate with Steven Strogatz at Cornell University. She received an Sc.B. in Applied Mathematics and an A.B. in Philosophy from Brown University. Her research is in the intersection of applied mathematics and atmospheric sciences. She builds and works with models of varying complexity: from energy balance models to global climate models. She is interested in using dynamical systems to explain the underlying mechanisms of atmosphere dynamics and model social phenomena.

Abstract: Voter models can provide insight into voter behavior and help forecast elections. Many popular voter models are based on opinion dynamics. However, patterns in voter turnout can affect election outcomes and distort representation of opinions. Furthermore, voter turnout varies highly depending on type, place, and size of elections. In our agent-based model, we take voter turnout into account by making a distinction between voter opinion and the act of voting. We assign an opinion to each node, which represents a potential voter. The potential voters make a decision to vote or to abstain based on their local network information. We explore a variety of network structures and node distributions and find that certain structures and parameter regimes favor undemocratic outcomes where a minority faction wins, especially when the local opinion distribution is not representative of the global opinion. We generalize the agent-based model to a community-based turnout model in order to gain insight into the network structure of US precinct dual graphs.

Website: https://www.kathlandgren.com/

Bio: Vandana Ravindran is a Scientia fellow at the Institute of Basic Medical Sciences. She received her Ph. D. from the Dhirubhai Ambani Institute of Information and Communication Technology, India and has since been a Royal Society-SERB Newton International fellow at the MRC-University of Glasgow Centre for Virus Research (2018-2020). Prior to her PhD, she worked as a research fellow at the Thrombosis Research Institute, India.

Abstract: Virus replication is entirely dependent on the host system they infect. This involves a high degree of virus-host specificity at the molecular level. For example, recognition of receptors on specific cell types by virus molecules is key to cell-entry, and interacting with host molecules is necessary to exploit intra-cellular 'machinery' to replicate. To achieve this, virus molecules must interact with many host molecules through a complex network of mostly protein-protein interactions (PPIs). In turn the host response to infection involves anti-viral factors, and subsequent virus response to host response, leading to a complex entanglement of virus and host interactions. In recent years control theory has been applied to biological systems with the aim of identifying the minimum set of molecular interactions that can drive the network to a required state. However, in an intra-cellular network it is unclear how control can be achieved in practice.

Viral infection is unique as a system for the study of the applicability of controllability to a natural system, as the virus makes many interactions with the host system and is explicitly exploiting host functions. To address this limitation we use viral infection, specifically human immunodeficiency virus type 1 (HIV-1), as a paradigm to model control of an infected cell. Our network controllability analysis demonstrates how a virus efficiently brings the dynamically organised host system into its control by mostly targeting existing critical control nodes, requiring fewer nodes than in the uninfected network. The lower number of control nodes is presumably to optimise exploitation of specific sub-systems needed for virus replication and/or involved in the host response to infection.

Website: https://www.med.uio.no/imb/english/people/aca/vandanar/

Bio: Allison Morgan is a computer science PhD student and NSF Graduate Research Fellow at the University of Colorado, Boulder. She received a B.A. in physics from Reed College. As a PhD student at UC Boulder, she is a member of Aaron Clauset's research group. Her interests lie within computational social science; more specifically, she measures the structural factors that drive a lack of diversity in science and highlight their consequences. Her research has been published in EPJ Data Science and PNAS, and covered by outlets such as the Washington Post and Scientific American.

Abstract: Across academia, men and women tend to publish at unequal rates. Existing explanations include the potentially unequal impact of parenthood on scholarship, but a lack of appropriate data has prevented its clear assessment. Here, we quantify the impact of parenthood on scholarship using an extensive survey of the timing of parenthood events, longitudinal publication data, and perceptions of research expectations among 3064 tenure-track faculty at 450 PhD-granting computer science, history, and business departments across the U.S. and Canada, along with data on institution specific parental leave policies. Parenthood explains the gender productivity gap by lowering the average short-term productivity of mothers, even as parents tend to be slightly more productive on average than non-parents. However, the size of the productivity penalty for mothers appears to have shrunk over time. Women report that paid parental leave and adequate childcare are important factors in their recruitment and retention, despite nearly 40% of institutions offering no paid parental leave. These results have broad implications for efforts to improve the inclusiveness of scholarship.

Website:  https://allisonmorgan.github.io/

Bio: Kimberly Glass is an assistant professor at Brigham and Women's Hospital, a teaching hospital of Harvard Medical School. Her research focuses on integrating multiple sources of omics data and on understanding how biological mechanisms and contexts affect structure of gene regulatory networks. Her research group's primary interests are in developing methods and building computational tools that merge an appreciation of network analysis, biological function and translational impact.

Abstract: Rapidly evolving Omics technologies are providing unprecedented amounts of data that can yield new insights into biological processes, allowing us to develop a more unified understanding of the development and progression of complex disease. In most cases a single gene or pathway cannot fully characterize a disease. Rather, disease-related changes often involve simultaneous alterations to the genome, epigenome, transcriptome, and proteome of the cell. Networks provide a powerful approach for identifying disease-related biological mechanisms by establishing a framework for integrating multiple types of Omics information. Importantly, complex shifts in the regulatory networks can provide important insights into the underlying mechanisms influencing disease state. Our group has developed a suite of methods that support: (1) effective integration of multi-omic data to reconstruct gene regulatory networks; (2) analysis of these networks to identify changes in disease state; and (3) modeling of context-specific networks in order to link regulatory alterations with specific phenotypes. In this talk, I will review how we have used these approaches to discover new features of disease and to understand the complex regulatory processes at work across individuals.

Website: https://sites.google.com/a/channing.harvard.edu/kimberlyglass/