CoChat

Returners and explorers dichotomy in human mobility

Open access
Pappalardo, Luca, Simini, Filippo, Rinzivillo, Salvatore, Pedreschi, Dino, Giannotti, Fosca, Barabási, Albert-László · 2015 · Nature Communications

The DOI matches its Crossref record; 45 of the 45 references it cites were found in a registry or an index.

Abstract

The availability of massive digital traces of human whereabouts has offered a series of novel insights on the quantitative patterns characterizing human mobility. In particular, numerous recent studies have lead to an unexpected consensus: the considerable variability in the characteristic travelled distance of individuals coexists with a high degree of predictability of their future locations. Here we shed light on this surprising coexistence by systematically investigating the impact of recurrent mobility on the characteristic distance travelled by individuals. Using both mobile phone and GPS data, we discover the existence of two distinct classes of individuals: returners and explorers. As existing models of human mobility cannot explain the existence of these two classes, we develop more realistic models able to capture the empirical findings. Finally, we show that returners and explorers play a distinct quantifiable role in spreading phenomena and that a correlation exists between their mobility patterns and social interactions.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/ncomms9166

Cites

45 of 45 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 A visualization of the complexity of the explored mobility patterns. A fragment of the GPS trajectories used in our study, displaying trips originating in the metropolitan areas of Pisa (in blue) and Florence (red). This plain geo-referenced visualization of experimental data reveals the confrontation of two ‘competing' metropolitan areas. It also demonstrates the ability of Big Data to portray social complexity. This map has been generated through the QGIS software, available at http://www.qgis.org/en/site/.
Figure 1 A visualization of the complexity of the explored mobility patterns. A fragment of the GPS trajectories used in our study, displaying trips originating in the metropolitan areas of Pisa (in blue) and Florence (red). This plain geo-referenced visualization of experimental data reveals the confrontation of two ‘competing' metropolitan areas. It also demonstrates the ability of Big Data to portray social complexity. This map has been generated through the QGIS software, available at http://www.qgis.org/en/site/.CC BY · PubMed Central
Figure 2 The distributions of k-radii and total radii. The distributions of total rg and , with k=2, 6, 10 for the GSM data (a–c) and GPS data (d–f). Black circles indicate the total rg, red and blue triangles indicate the for the GSM and GPS data, respectively. All distributions are approximated by a truncated power law ; the dashed black line represents a truncated power-law fit of the total rg, the red and blue solid lines represent a truncated power-law fit of for the GSM and GPS data, respectively. Table 1 shows the fitting parameters of the truncated power laws.
Figure 2 The distributions of k-radii and total radii. The distributions of total rg and , with k=2, 6, 10 for the GSM data (a–c) and GPS data (d–f). Black circles indicate the total rg, red and blue triangles indicate the for the GSM and GPS data, respectively. All distributions are approximated by a truncated power law ; the dashed black line represents a truncated power-law fit of the total rg, the red and blue solid lines represent a truncated power-law fit of for the GSM and GPS data, respectively. Table 1 shows the fitting parameters of the truncated power laws.CC BY · PubMed Central
Figure 3 The correlation between recurrent and overall mobility. The scatter plots represent the correlation between total rg and for k=2, 4, 8 in the GSM data set (a–c) and the GPS data set (d–f). Each point is coloured from blue to red, indicating the density of points in the corresponding region. Most of the points gather around the x-axis, the diagonal and the origin. The insets magnify the origin of the plot to [0, 100 km] for GSM and [0, 16 km] for GPS, demonstrating that the split emerges for smaller radii as well. As k increases explorers become returners. This transition is faster in the GPS case, consistent with the fact that the vehicle mobility represents a subset of trips and visited locations.
Figure 3 The correlation between recurrent and overall mobility. The scatter plots represent the correlation between total rg and for k=2, 4, 8 in the GSM data set (a–c) and the GPS data set (d–f). Each point is coloured from blue to red, indicating the density of points in the corresponding region. Most of the points gather around the x-axis, the diagonal and the origin. The insets magnify the origin of the plot to [0, 100 km] for GSM and [0, 16 km] for GPS, demonstrating that the split emerges for smaller radii as well. As k increases explorers become returners. This transition is faster in the GPS case, consistent with the fact that the vehicle mobility represents a subset of trips and visited locations.CC BY · PubMed Central
Figure 4 The ratio between recurrent and overall mobility. The distribution P(sk) of the ratio measured on the GSM data for k=2, 4, 8 (a–c). The peak at sk=0 corresponds to explorers, while the sk=1 peak corresponds to returners. For small k in the GSM data, k-explorers are more numerous than k-returners. As k increases the number of k-returners increases and overcomes the number of k-explorers. A balance in the population is reached at k=4. (d–f) The P(sk) for the GPS data. We again observe two peaks, but the k-returners peak, sk=1, dominates for all k≥2.
Figure 4 The ratio between recurrent and overall mobility. The distribution P(sk) of the ratio measured on the GSM data for k=2, 4, 8 (a–c). The peak at sk=0 corresponds to explorers, while the sk=1 peak corresponds to returners. For small k in the GSM data, k-explorers are more numerous than k-returners. As k increases the number of k-returners increases and overcomes the number of k-explorers. A balance in the population is reached at k=4. (d–f) The P(sk) for the GPS data. We again observe two peaks, but the k-returners peak, sk=1, dominates for all k≥2.CC BY · PubMed Central
Figure 5 The individual mobility networks of returners and explorers. The mobility networks of returners and explorers for k=2. Nodes (circles) indicate the geographic locations visited by the individual, and each link denotes a travel observed between two locations. When the total rg is small, the two most important locations (red and blue) are close to each other for both two-explorers and two-returners. As the total radius increases the behaviour of two-returners and two-explorers starts to differ; for two-returners, the two most important locations move away from each other; for two-explorers, they stay close and other clusters of locations emerge far from the centre of mass (the grey cross).
Figure 5 The individual mobility networks of returners and explorers. The mobility networks of returners and explorers for k=2. Nodes (circles) indicate the geographic locations visited by the individual, and each link denotes a travel observed between two locations. When the total rg is small, the two most important locations (red and blue) are close to each other for both two-explorers and two-returners. As the total radius increases the behaviour of two-returners and two-explorers starts to differ; for two-returners, the two most important locations move away from each other; for two-explorers, they stay close and other clusters of locations emerge far from the centre of mass (the grey cross).CC BY · PubMed Central
Figure 6 EPR model predictions. (a,b) The prediction of the EPR model for k=2. We find that two-explorers dominate the population of synthetic individuals and the balance in the population is reached only for k=60, in contrast with k=4 in the empirical data. (d,e) The results of the d-EPR model for k=2. In this case, the two-explorers continue to dominate the population, although the balance is reached at lower values of k=9, coming closer to empirical data. The insets in a,d magnify the plot at smaller values of the radii of gyration. Plots (c,f) show how the number of k-returners and k-explorers changes with k for EPR model and d-EPR model, respectively.
Figure 6 EPR model predictions. (a,b) The prediction of the EPR model for k=2. We find that two-explorers dominate the population of synthetic individuals and the balance in the population is reached only for k=60, in contrast with k=4 in the empirical data. (d,e) The results of the d-EPR model for k=2. In this case, the two-explorers continue to dominate the population, although the balance is reached at lower values of k=9, coming closer to empirical data. The insets in a,d magnify the plot at smaller values of the radii of gyration. Plots (c,f) show how the number of k-returners and k-explorers changes with k for EPR model and d-EPR model, respectively.CC BY · PubMed Central
Figure 7 Best friend interaction patterns of returners and explorers. The histograms indicate the distributions of the fraction of two-returners whose best friend (that is, the most called contact) is a two-returner (RR) (a) or a two-explorer (RE) (b), and two-explorers whose best friend is a two-returner (ER) (c) or two-explorer (EE) (d), obtained from 100,000 randomized experiments where we randomly reassign each individual's best friend. The dashed line indicates the real fraction of two-returners (two-explorers) whose best friend is a two-returner (two-explorer). We observe that individuals that communicate with each other are more likely to belong to the same mobility group than by chance.
Figure 7 Best friend interaction patterns of returners and explorers. The histograms indicate the distributions of the fraction of two-returners whose best friend (that is, the most called contact) is a two-returner (RR) (a) or a two-explorer (RE) (b), and two-explorers whose best friend is a two-returner (ER) (c) or two-explorer (EE) (d), obtained from 100,000 randomized experiments where we randomly reassign each individual's best friend. The dashed line indicates the real fraction of two-returners (two-explorers) whose best friend is a two-returner (two-explorer). We observe that individuals that communicate with each other are more likely to belong to the same mobility group than by chance.CC BY · PubMed Central
Figure 8 Interaction patterns of returners and explorers versus the rank of friendship. (a) The fraction of two-returners having a two-returner (RR) or a two-explorer (RE) as n-th best friend as n increases. (b) The fraction of two-explorers having a two-returner (ER) or a two-explorer (EE) as n-th best friend as n increases. We observe that the observed the fractions are significantly higher (RR and EE) or significantly lower (ER and RE) than those obtained by chance for all n up to 15.
Figure 8 Interaction patterns of returners and explorers versus the rank of friendship. (a) The fraction of two-returners having a two-returner (RR) or a two-explorer (RE) as n-th best friend as n increases. (b) The fraction of two-explorers having a two-returner (ER) or a two-explorer (EE) as n-th best friend as n increases. We observe that the observed the fractions are significantly higher (RR and EE) or significantly lower (ER and RE) than those obtained by chance for all n up to 15.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 45 references

Research works printed but not linked: 4

  1. Giannotti, F., Nanni, M., Pinelli, F. & Pedreschi, D. in Proceedings of the 13th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD) 330–339ACM (2007).
  2. Jiang, S. et al. in Proceedings of the 2nd ACM SIGKDD (International Workshop on Urban Computing (UrbComp'13)) 21–29ACM (2013).
  3. Rinzivillo, S. et al. in Proceedings of the 1st International Conference on Data Science and Advanced Analytics (DSAA'2014) IEEE Computer Society, Shanghai, China (2014).
  4. Wang, D., Pedreschi, D., Chaoming, S., Giannotti, F. & Barabási, A. L. in Proceedings of the 17th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD) 1100–1108ACM New York, San Diego, CA, USA (2011).

Cited by

See all 494

Metadata

Authors
Pappalardo, L., Simini, F., Rinzivillo, S., Pedreschi, D., Giannotti, F., & Barabási, A.-L.
Affiliations
  • University of Pisa (IT)
  • Northeastern University (US)
  • Istituto di Scienza e Tecnologie dell'Informazione "Alessandro Faedo" (IT)
Corresponding author
  • Pappalardo, Luca
  • Simini, Filippo
Venue
Nature Communications
Published
2015-09-08
Volume
6
Issue
1
Pages
8166-8166
ISSN
2041-1723 (ISSN-L)
OpenAlex
W2134946452
Cited by
494
References
45
Citations by year
2026: 43 · 2025: 42 · 2024: 47 · 2023: 52 · 2022: 49 · 2021: 72 · 2020: 55 · 2019: 42 · 2018: 33 · 2017: 35 · 2016: 22 · 2015: 1
FWCI
57.10
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Communications (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Bristol Research (University of Bristol) (publishedVersion, other-oa)
  • Digital Access to Scholarship at Harvard (DASH) (Harvard University) (submittedVersion, cc-by)
  • RePEc: Research Papers in Economics (submittedVersion)
  • CINECA IRIS Institutial research information system (University of Pisa) (submittedVersion, cc-by)
  • ISTI Open Portal (submittedVersion, other-oa)
  • Open Research (Auckland University of Technology) (submittedVersion, cc-by)
  • Explore Bristol Research (submittedVersion, other-oa)
Funders
  • European Commission (H2020 Program, FP7 ICT, 270833, H2020, 317532, 641191)
  • Defense Threat Reduction Agency (HDTRA1)
  • Horizon 2020 Framework Programme (641191)
  • FP7 Information and Communication Technologies
Topics
  • Social Sciences › Transportation › Human Mobility and Location-Based Analysis
  • Physical Sciences › Physics and Astronomy › Statistical and Nonlinear Physics › Complex Network Analysis Techniques
  • Physical Sciences › Computer Science › Computer Networks and Communications › Opportunistic and Delay-Tolerant Networks
Keywords
human mobility, mobile phone data, GPS data, mobility models, epidemic spread, social interaction
MeSH terms
Exploratory Behavior, Humans, Interpersonal Relations, Models, Theoretical, Social Behavior, Travel, Geographic Information Systems, Cell Phone

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Returners and explorers dichotomy in human mobility. https://app.cochat.ai/papers/10.1038/ncomms9166
The paper
Pappalardo, L., Simini, F., Rinzivillo, S., Pedreschi, D., Giannotti, F., & Barabási, A.-L. (2015). Returners and explorers dichotomy in human mobility. Nature Communications. https://doi.org/10.1038/ncomms9166

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.