Temporal fidelity in dynamic social networks

Abstract

It has recently become possible to record detailed social interactions in large social systems with high resolution. As we study these datasets, human social interactions display patterns that emerge at multiple time scales, from minutes to months. On a fundamental level, understanding of the network dynamics can be used to inform the process of measuring social networks. The details of measurement are of particular importance when considering dynamic processes where minute-to-minute details are important, because collection of physical proximity interactions with high temporal resolution is difficult and expensive. Here, we consider the dynamic network of proximity-interactions between approximately 500 individuals participating in the Copenhagen Networks Study. We show that in order to accurately model spreading processes in the network, the dynamic processes that occur on the order of minutes are essential and must be included in the analysis. IntroductionTemporal networks provide an important framework for modeling a variety of real systems 1 .Examples of complex systems where dynamics can play a central role include social networks, energy grids, networks of sexual contacts, and transportation systems 2-8 .Only recently, thanks to technical developments in data collection, it has become possible to collect high-resolution data about physical and virtual interactions in complex social systems.Using sociometric badges or smartphones, it is now possible to record interactions happening 1 arXiv:1507.01484v2 [physics.soc-ph] 21 Aug 2015 on multiple channels and at multiple timescales, measuring events with minute-by-minute resolution 6,[9][10][11][12] . With access to such data, we can begin to describe the complexity, structure, and dynamics of such social systems 13 . Accurate measurements and models of social systems are necessary in order to understand how diseases spread 6,7 , what makes teams productive 14, 15 , or how friendships form and disappear 11,13 .A fully-formed framework for incorporating network dynamics has yet to be established ?, ?, 16-18 .We know, however, that for many practical applications, it is important to get the details right, because variations in how the time dimension is incorporated can lead to significant differences in the modeling of dynamical processes unfolding on the network. Understanding spreading in dynamic networks is of particular interest, as these may represent a wide variety of processes in the system, including spreading of biological pathogens, information, knowledge, or behaviors.Recently, there has been a growing interest in how to correctly and efficiently incorporate time dimension in the modeling of disease spread. Over the last few years, studies have focused on the mixing matrices capturing important epidemiological features ? , efficient representation of the spreading networks with coarser temporal representation ?, 16 , and the fundamental impact that temporal features have on the spreading process 17, 18 . Here we study how the fidelity of representation...

Modern temporal network theory: a colloquium

“…Typically one filters out very brief contacts, so the remaining contacts capture people either in a face-to-face conversation or in a crowd or gathering of some kind. WiFi [255,293,112] or Bluetooth [277,263,276,225,262] sensors give almost the same resolution (with a few meters lower positional accuracy but worse angular resolution). Of these references, Refs.…”

Interaction data from the Copenhagen Networks Study

“…For more details on how aggregation changes the structure of the network and the subsequent implications for epidemic modelling, refer to Stopczynski et al . 43 . For a more detailed description of the network structures at the highest resolution, and the implications for modeling of social interactions, refer to Sekara et al .…”

Optimizing targeted vaccination across cyber–physical networks: an empirically based mathematical simulation study

“…In other words, if there is a link in any 5 min bin throughout the period of interest, we add a 5 min link to the simulation extending an (possibly shorter) interaction to 5 min. The temporal resolution of 5 min has been shown to capture the key dynamics of person-to-person networks [18,30]. We stress that while sensorbased proximity networks are a highly useful model to assess the impact of immunizations strategies, they suffer from important limitations [31][32][33]; see Discussion for details.…”