There is a contradiction at the heart of our current understanding of individual and collective mobility patterns. On one hand, a highly influential stream of literature on human mobility driven by analyses of massive empirical datasets finds that hu
man movements show no evidence of characteristic spatial scales. There, human mobility is described as scale-free. On the other hand, in geography, the concept of scale, referring to meaningful levels of description from individual buildings through neighborhoods, cities, regions, and countries, is central for the description of various aspects of human behavior such as socio-economic interactions, or political and cultural dynamics. Here, we resolve this apparent paradox by showing that day-to-day human mobility does indeed contain meaningful scales, corresponding to spatial containers restricting mobility behavior. The scale-free results arise from aggregating displacements across containers. We present a simple model, which given a persons trajectory, infers their neighborhoods, cities, and so on, as well as the sizes of these geographical containers. We find that the containers characterizing the trajectories of more than 700,000 individuals do indeed have typical sizes. We show that our model generates highly realistic trajectories without overfitting and provides a new lens through which to understand the differences in mobility behaviour across countries, gender groups, and urban-rural areas.
The global public sphere has changed dramatically over the past decades: a significant part of public discourse now takes place on algorithmically driven platforms owned by a handful of private companies. Despite its growing importance, there is scan
t large-scale academic research on the long-term evolution of user behaviour on these platforms, because the data are often proprietary to the platforms. Here, we evaluate the individual behaviour of 600,000 Twitter users between 2012 and 2019 and find empirical evidence for an acceleration of the way Twitter is used on an individual level. This manifests itself in the fact that cohorts of Twitter users behave differently depending on when they joined the platform. Behaviour within a cohort is relatively consistent over time and characterised by strong internal interactions, but over time behaviour from cohort to cohort shifts towards increased activity. Specifically, we measure this in terms of more tweets per user over time, denser interactions with others via retweets, and shorter content horizons, expressed as an individuals decaying autocorrelation of topics over time. Our observations are explained by a growing proportion of active users who not only tweet more actively but also elicit more retweets. These behaviours suggest a collective contribution to an increased flow of information through each cohorts news feed -- an increase that potentially depletes available collective attention over time. Our findings complement recent, empirical work on social acceleration, which has been largely agnostic about individual user activity.
The full range of activity in a temporal network is captured in its edge activity data -- time series encoding the tie strengths or on-off dynamics of each edge in the network. However, in many practical applications, edge-level data are unavailable,
and the network analyses must rely instead on node activity data which aggregates the edge-activity data and thus is less informative. This raises the question: Is it possible to use the static network to recover the richer edge activities from the node activities? Here we show that recovery is possible, often with a surprising degree of accuracy given how much information is lost, and that the recovered data are useful for subsequent network analysis tasks. Recovery is more difficult when network density increases, either topologically or dynamically, but exploiting dynamical and topological sparsity enables effective solutions to the recovery problem. We formally characterize the difficulty of the recovery problem both theoretically and empirically, proving the conditions under which recovery errors can be bounded and showing that, even when these conditions are not met, good quality solutions can still be derived. Effective recovery carries both promise and peril, as it enables deeper scientific study of complex systems but in the context of social systems also raises privacy concerns when social information can be aggregated across multiple data sources.
The ability to share social network data at the level of individual connections is beneficial to science: not only for reproducing results, but also for researchers who may wish to use it for purposes not foreseen by the data releaser. Sharing such d
ata, however, can lead to serious privacy issues, because individuals could be re-identified, not only based on possible nodes attributes, but also from the structure of the network around them. The risk associated with re-identification can be measured and it is more serious in some networks than in others. Various optimization algorithms have been proposed to anonymize the network while keeping the number of changes minimal. However, existing algorithms do not provide guarantees on where the changes will be made, making it difficult to quantify their effect on various measures. Using network models and real data, we show that the average degree of networks is a crucial parameter for the severity of re-identification risk from nodes neighborhoods. Dense networks are more at risk, and, apart from a small band of average degree values, either almost all nodes are re-identifiable or they are all safe. Our results allow researchers to assess the privacy risk based on a small number of network statistics which are available even before the data is collected. As a rule-of-thumb, the privacy risks are high if the average degree is above 10. Guided by these results we propose a simple method based on edge sampling to mitigate the re-identification risk of nodes. Our method can be implemented already at the data collection phase. Its effect on various network measures can be estimated and corrected using sampling theory. These properties are in contrast with previous methods arbitrarily biasing the data. In this sense, our work could help in sharing network data in a statistically tractable way.
Language can exert a strong influence on human behaviour. In experimental studies, it is for example well-known that the framing of an experiment or priming at the beginning of an experiment can alter participants behaviour. However, few studies have
been conducted to determine why framing or priming specific words can alter peoples behaviour. Here, we show that the behaviour of participants in a game-theoretical experiment is driven mainly by social norms, and that participants adherence to different social norms is influenced by the exposure to economic terminology. To explore how these terminology-driven changes impact behavior at the system level, we use established frameworks for modeling collective cooperative behaviour. We find that economic terminology induces a behavioural difference which is larger than that caused by financial incentives in the magnitude usually employed in experiments and simulation. These findings place an increased responsibility on scientists and science communicators, as scientific terminology is increasingly communicated to the general population.
According to personality psychology, personality traits determine many aspects of human behaviour. However, validating this insight in large groups has been challenging so far, due to the scarcity of multi-channel data. Here, we focus on the relation
ship between mobility and social behaviour by analysing trajectories and mobile phone interactions of $sim 1,000$ individuals from two high-resolution longitudinal datasets. We identify a connection between the way in which individuals explore new resources and exploit known assets in the social and spatial spheres. We show that different individuals balance the exploration-exploitation trade-off in different ways and we explain part of the variability in the data by the big five personality traits. We point out that, in both realms, extraversion correlates with the attitude towards exploration and routine diversity, while neuroticism and openness account for the tendency to evolve routine over long time-scales. We find no evidence for the existence of classes of individuals across the spatio-social domains. Our results bridge the fields of human geography, sociology and personality psychology and can help improve current models of mobility and tie formation.
Humans interact through numerous channels to build and maintain social connections: they meet face-to-face, initiate phone calls or send text messages, and interact via social media. Although it is known that the network of physical contacts, for exa
mple, is distinct from the network arising from communication events via phone calls and instant messages, the extent to which these networks differ is not clear. In fact, the network structure of these channels shows large structural variations. Each network of interactions, however, contains both central and peripheral individuals: central members are characterized by higher connectivity and can reach a high fraction of the network within a low number of connections, contrary to the nodes on the periphery. Here we show that the various channels account for diverse relationships between pairs of individuals and the corresponding interaction patterns across channels differ to an extent that hinders the simple reduction of social ties to a single layer. Furthemore, the origin and purpose of each network also determine the role of their respective central members: highly connected individuals in the person-to-person networks interact with their environment in a regular manner, while members central in the social communication networks display irregular behavior with respect to their physical contacts and are more active through rare, social events. These results suggest that due to the inherently different functions of communication channels, each one favors different social behaviors and different strategies for interacting with the environment. Our findings can facilitate the understanding of the varying roles and impact individuals have on the population, which can further shed light on the prediction and prevention of epidemic outbreaks, or information propagation.
The recent availability of digital traces generated by phone calls and online logins has significantly increased the scientific understanding of human mobility. Until now, however, limited data resolution and coverage have hindered a coherent descrip
tion of human displacements across different spatial and temporal scales. Here, we characterise mobility behaviour across several orders of magnitude by analysing ~850 individuals digital traces sampled every ~16 seconds for 25 months with ~10 meters spatial resolution. We show that the distributions of distances and waiting times between consecutive locations are best described by log-normal distributions and that natural time-scales emerge from the regularity of human mobility. We point out that log-normal distributions also characterise the patterns of discovery of new places, implying that they are not a simple consequence of the routine of modern life.
Predictive models for human mobility have important applications in many fields such as traffic control, ubiquitous computing and contextual advertisement. The predictive performance of models in literature varies quite broadly, from as high as 93% t
o as low as under 40%. In this work we investigate which factors influence the accuracy of next-place prediction, using a high-precision location dataset of more than 400 users for periods between 3 months and one year. We show that it is easier to achieve high accuracy when predicting the time-bin location than when predicting the next place. Moreover we demonstrate how the temporal and spatial resolution of the data can have strong influence on the accuracy of prediction. Finally we uncover that the exploration of new locations is an important factor in human mobility, and we measure that on average 20-25% of transitions are to new places, and approx. 70% of locations are visited only once. We discuss how these mechanisms are important factors limiting our ability to predict human mobility.
We introduce geoplotlib, an open-source python toolbox for visualizing geographical data. geoplotlib supports the development of hardware-accelerated interactive visualizations in pure python, and provides implementations of dot maps, kernel density
estimation, spatial graphs, Voronoi tesselation, shapefiles and many more common spatial visualizations. We describe geoplotlib design, functionalities and use cases.
