Quantifying agent impacts on contact sequences in social interactions

Abstract

Human social behavior plays a crucial role in how pathogens like SARS-CoV-2 or fake news spread in a population. Social interactions determine the contact network among individuals, while spreading, requiring individual-to-individual transmission, takes place on top of the network. Studying the topological aspects of a contact network, therefore, not only has the potential of leading to valuable insights into how the behavior of individuals impacts spreading phenomena, but it may also open up possibilities for devising effective behavioral interventions. Because of the temporal nature of interactions—since the topology of the network, containing who is in contact with whom, when, for how long, and in which precise sequence, varies (rapidly) in time—analyzing them requires developing network methods and metrics that respect temporal variability, in contrast to those developed for static (i.e., time-invariant) networks. Here, by means of event mapping, we propose a method to quantify how quickly agents mingle by transforming temporal network data of agent contacts. We define a novel measure called contact sequence centrality, which quantifies the impact of an individual on the contact sequences, reflecting the individual’s behavioral potential for spreading. Comparing contact sequence centrality across agents allows for ranking the impact of agents and identifying potential ‘behavioral super-spreaders’. The method is applied to social interaction data collected at an art fair in Amsterdam. We relate the measure to the existing network metrics, both temporal and static, and find that (mostly at longer time scales) traditional metrics lose their resemblance to contact sequence centrality. Our work highlights the importance of accounting for the sequential nature of contacts when analyzing social interactions.

Hidden dependence of spreading vulnerability on topological complexity

“…It is for this reason that S(t, t ), in comparison to S (t, t ), is not simply capturing the higher order structures in a temporal network, but is actually respecting the correct sequential information of agent interactions in the topology of the network. This issue has recently been discussed more extensively in a recent paper coauthored by three of us [29].…”

The Lighting of the BECONs

“…In addition, such thought experiments suggest that the question under which conditions which BECONs can adequately track virus transmission is open for future research; one could imagine, for instance, implementing different interventions and running epidemiological virus spread models on the resulting networks to understand how different interventions change the epidemiological course of the virus. Finally, the current setup ignores dynamical information about interventions (e.g., the duration of effects), and extending measurements and models in this direction could augment the signal considerably (Dekker et al, 2021). Such information could then be used to assess these interventions in a more precise fashion.…”

“…For example, in the case of COVID-19, the presence of aerosol transmission or infection via surfaces can create links between people who are at a greater distance than 1.5 meters, or between people who were present at the same place at distinct time points (e.g., because the aerosols remain present in bathrooms after the infectious person has left). Second, if the network integrates contacts over time (e.g., by taking the union of all contact networks at each time point, which registers during a certain time interval who has ever been in contact with who, but not when), the representation will contain false positive connections; for instance, when A and B were in contact, and subsequently B and C were in contact, then the patterns of links suggests that both A → B → C and C → B → A are possible infection routes, while only the former route is possible (see also Dekker et al (2021)). Third, various kinds of measurement errors can yield false positives and false negatives.…”

“…Thus, we assume that the time-dependent network has similar properties as the properties of the class and thus abstract the temporal changes in time. Finally, novel approaches based on the analysis of time series data can be used to include the dynamic changes of the network in the analysis (Dekker et al, 2021). In the current paper, we focus on the simplest case, namely one in which the contact network is stable over time so that it can be characterized by a single network structure.…”