Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

The NetMob25 Dataset: A High-resolution Multi-layered View of Individual Mobility in Greater Paris Region

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper introduces the NetMob25 dataset: seven days of GPS traces from 3,337 Greater Paris residents, with about 500 million points and over 80,000 validated trips, plus census-calibrated weights.

desk verdict A potentially valuable GPS mobility dataset, but the abstract's ~500M point count is off by an order of magnitude from the trip statistics; needs major revision before it can be trusted. read the letter →

arxiv 2506.05903 v1 pith:3KBVPX3K submitted 2025-06-06 cs.CY cs.IR

Alexandre Chasse , Anne J. Kouam , Aline C. Viana , Razvan Stanica , Wellington V. Lobato , Geymerson Ramos , Geoffrey Deperle , Abdelmounaim Bouroudi
show 2 more authors
Suzanne Bussod Fernando Molano
This is my paper
classification cs.CYcs.IR
keywords GPSmobilitysurveydatasetGreaterParistravelbehaviortransportmodestrippurposesstatisticalweightinganonymization
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

NetMob25 is a GPS-based mobility dataset covering a full week for 3,337 volunteer residents of the Greater Paris region, collected between October 2022 and May 2023. The paper's claim is that this dataset gives researchers a rare ground-truth view of individual daily mobility: roughly 500 million location points sampled every 2–3 seconds, over 80,000 trips whose transport modes and purposes were validated in phone interviews, and linked sociodemographic and household profiles. Because the survey combines passive GPS logging with a participant logbook and human verification, the authors argue the trip annotations are behaviorally reliable. Calibrated statistical weights are included so the sample can be extrapolated to the regional population. The paper is the reference description of the survey design, processing, and anonymization that the released data depends on.

What carries the argument

The carrying mechanism is the hybrid survey protocol: a BT-Q1000XT GPS travel recorder configured to log a position every 2–3 seconds only while motion is detected, a daily travel diary kept by each participant, and a follow-up phone interview in which inferred trips, modes, and purposes are verified or corrected. This protocol produces the trip-level ground truth that makes the raw traces interpretable. Around it sit three supporting machinery pieces: an H3 hexagonal grid blurring step that hides the first and last 50–100 meters of each trip, an anonymization and cleaning pipeline that removes non-trip points, and census-calibrated individual and trip weights that allow extrapolation to the Greater Paris population.

What would settle it

Take a subsample of participants and have them carry a second, independent always-on GPS logger (or share their phone's location history) for the same week; if the second logger records trips that are absent from the validated NetMob25 traces, or records movement during periods the device shows as a gap, the claim that the traces form a complete weekly mobility record would be contradicted.

Watch

Extended reading notes

Core claim

The central claim is that the dataset, built from the EMG 2023 GNSS-based mobility survey, provides a high-resolution, multi-layered record of individual mobility in a large European metropolis. It combines three linkable databases — individual sociodemographic and household characteristics, validated trips with timestamps, modes, and purposes, and raw GPS traces with points every 2–3 seconds — so that trajectories, behaviors, and population profiles can be analyzed together. The authors further claim that the collected sample, once reweighted against census marginals, supports population-level estimates of trip volumes, mode use, and spatial flows, and that the anonymization pipeline preserves analytical value by blurring only trip endpoints while keeping in-trip points at full resolution.

Load-bearing premise

The load-bearing premise is that participants carried the BT-Q1000XT device with them for seven consecutive days and that the device recorded a position whenever they moved, so that every gap in the GPS trace means a stationary period or no travel rather than a missed or unlogged trip.

Editorial extensions

If this is right

  • Researchers can estimate population-level mobility for Greater Paris by applying the provided individual and trip weights to trip counts, mode shares, and spatial flows.
  • The 2–3-second GPS traces, time-aligned with validated trips, can serve as reference data for evaluating trip detection and segmentation algorithms.
  • Linked sociodemographic and trip data allow studies of transport mode choice, multimodal chains, teleworking patterns, and mobility differences across age, sex, and household type.
  • Annotated special days and 'No Trip' or 'No Traces' rows let analysts filter or study atypical mobility during holidays, strikes, weekends, and school breaks.
  • The anonymization design preserves in-trip trajectory detail while obscuring home and work locations, supporting spatial analyses of flows at fine geographic scale.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the motion-triggered logging behaved as described, then stationary gaps can be treated as periods of no travel; however, the dataset cannot by itself distinguish 'device left at home' from 'no travel,' so studies of immobility should treat those two cases as potentially confounded.
  • The weight calibration is based on sociodemographic marginals and day of week, not on joint day-to-day mobility patterns; analyses of weekly rhythms or longitudinal dependencies may need to assess the extra uncertainty introduced by this.
  • The full-resolution in-trip points, with endpoints blurred independently per trip, create an opportunity to benchmark privacy-preserving trajectory publishing: a researcher could attempt to re-identify homes or workplaces from the blurred endpoints and probe the actual protection offered by the chosen cell resolution.
  • Because GPS files exist for only 3,320 of the 3,337 participants, any analysis that joins GPS traces to trip records should first verify whether the missing 17 introduce bias; the paper flags them but does not establish that their absence is random.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper describes the NetMob25 dataset, derived from the EMG 2023 GNSS-based mobility survey in the Île-de-France region. The dataset comprises three linked components: an individuals database with sociodemographic and household attributes for 3,337 participants, a trips database with over 80,000 validated trips annotated with modes and purposes, and a raw GPS traces database claimed to contain approximately 500 million high-frequency points. The paper presents the survey design, data collection protocol, anonymization pipeline, weighting methodology, and an exploratory characterization of individual and trip-level statistics, transport modes, and spatial flows. It also discusses intended use cases and access conditions under the NetMob25 Data Challenge.

Significance. If the dataset is as described, it would be a valuable resource for mobility research: it combines high-frequency GPS trajectories with human-validated trip boundaries, transport modes, purposes, and rich sociodemographic covariates, and it provides calibration weights for population-level inference. The multi-stage validation through logbooks and phone interviews is a notable strength, as is the careful documentation of anonymization steps (pseudonymization, removal of non-trip data, and H3-based blurring of trip endpoints). The exploratory analyses provide useful initial characterizations of travel behavior. However, the paper's current value as a dataset description is substantially weakened by internal inconsistencies in the headline-scale claims and the collection window, which must be resolved before the description can be trusted.

major comments (4)
  1. [Abstract; Section 3; Section 5.2] The claim of "approximately 500 million" GPS points is arithmetically inconsistent with the reported trip statistics. With 80,697 validated trips, a mean trip duration of 29.97 minutes (Fig. 8), and a 2–3 s sampling interval, the expected number of in-trip points is about 50–80 million (80,697 × 29.97 × 60 / 2.5 ≈ 58 million), roughly an order of magnitude below 500 million. Since Section 3 states that all non-trip points are discarded and Section 2 states that the device logs only during movement, the 500 million count implies either an average trip duration of about 4.3 hours, a sub-second sampling rate, or the inclusion of non-trip points—each contradicting the stated protocol. Please reconcile this discrepancy and specify the exact point count in the released GPS database.
  2. [Section 5.2 vs. Section 2 and Abstract] The data collection window is described inconsistently: Section 2 and the abstract state that data were collected between mid-October 2022 and mid-May 2023 (20 weeks), while Section 5.2 states that the dataset covers October 17, 2022 to January 15, 2023. Moreover, Fig. 5(a) shows trip activity between 2023-03-25 and 2023-04-15, which falls outside the January 15 end date. This contradiction affects all temporal-coverage claims and any analyses of daily, weekly, or seasonal patterns. The actual dates present in the released data must be clarified.
  3. [Section 4.4 and Section 5.1] The representativity evaluation in Section 5.1 is partially circular. The individual weights constructed in Section 4.4 are calibrated to INSEE census margins (department, age, sex, socio-professional category, household size, etc.), and the weighted age and sex distributions in Section 5.1 are then compared to those same INSEE margins. The resulting "good alignment" is a direct consequence of the calibration procedure, not an independent validation of sample representativity. The chi-squared test on raw counts (Fig. 3) is a valid check of the unweighted sample, but the paper should distinguish raw-sample goodness-of-fit from weighted alignment and ideally validate the weights on variables or margins not used in calibration (e.g., against the EGT survey or withheld INSEE variables).
  4. [Section 2] The interpretation of GPS gaps as stationary periods or absence of mobility relies on the assumption that participants carried the BT-Q1000XT continuously for seven days and that the motion-triggered logging reliably captures short or slow trips. The paper does not report any compliance statistics (e.g., the fraction of participants with valid recordings on all seven days, the distribution of daily recording durations) or a validation of the device's motion-trigger behavior against the logbook data. Without such information, the completeness of individual daily trajectories and the interpretation of "No Trip" days are difficult to assess. Please add any available compliance checks or explicitly discuss this limitation in the data collection description.
minor comments (5)
  1. [Abstract] The phrase "an unique GPS-based mobility dataset" should be corrected to "a unique GPS-based mobility dataset".
  2. [Figure 5] The caption says "estimated weighted number of trips per date and per hour," but panel (a) is aggregated per date and panel (b) per hour; clarify this in the caption to avoid confusion.
  3. [Table 3] The orientation of the inter-departmental flow matrix (rows as origin or destination) should be stated explicitly in the caption; currently the reader must infer the meaning from the percentages and diagonal dominance.
  4. [Section 4.4] The description of the weight calibration would benefit from the exact calibration formula or a direct reference to the documentation on how the pairwise marginal distributions are combined, since readers may want to replicate or evaluate the weighting procedure.
  5. [Figure 9] The mode labels in the figure panels appear truncated (e.g., "ELECT_BIKE"); ensure the final version uses complete and readable mode names, and define abbreviations such as PRIV_CAR_DRIVER.

Circularity Check

1 steps flagged · score 2.0 of 10

Post-stratification alignment with INSEE is guaranteed by construction; no load-bearing circularity in the core dataset description.

  1. fitted input called prediction [Section 4.4 (Weight construction) and Section 5.1 (Individual-Level Statistics, Fig. 4).]
    "To ensure representativeness, calibration weights were computed independently for individuals and trips using pairwise marginal distributions derived from census statistics. As a corrective, post-stratification weights are applied to each individual to restore representativity (Fig. 4)."

    The individual weights are calibrated so their weighted margins match INSEE census counts for department, age, sex, socio-professional category, household size, car ownership, and diploma. The 'good alignment' with INSEE shown after weighting in Section 5.1 is therefore the calibration constraint restated, not an independent empirical validation. The paper presents this weighted alignment as evidence of representativity, but it is forced by the construction of the weights. The unweighted chi-square test against INSEE (p = 1.2e-9) is an independent demonstration of the original sample imbalance, so the tautology affects only the post-weighting alignment claim.

full rationale

This is a dataset description paper rather than a derivation-based or predictive paper, so most circularity patterns do not apply. The only reduction-by-construction found is the representativity check: weights are calibrated to INSEE margins in Section 4.4, then Section 5.1 presents the resulting weighted INSEE alignment as a positive characterization. That alignment is guaranteed by the calibration, but it is not the central claim and does not affect the released GPS, trip, or individual data. The paper's notable internal inconsistencies, for example the roughly 500 million claimed GPS points versus about 80 thousand validated trips at 2-3 second sampling and the conflicting collection-window dates in Section 5.2 versus the abstract, are quantitative correctness or data-quality issues rather than circularity and are not scored here. No load-bearing self-citations, imported uniqueness theorems, or ansatz-smuggling steps are present. Score 2 reflects a minor tautological validation claim within an otherwise self-contained dataset description.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

No invented entities are introduced. The central assumptions are about device behavior, validation reliability, weighting, and privacy. The weights are calibrated quantities, not free fitting parameters, so the free parameter list is empty. The main caveat is that several assumptions are untested and the data are NDA-gated.

assumptions (5)
  • domain assumption The BT-Q1000XT device logs positions only when motion is detected, so gaps in the recorded traces correspond to stationary periods or absence of mobility, not signal loss.
    Invoked in Section 2 to interpret data gaps as non-movement; the provided citation is the device manual, with no empirical validation in this paper.
  • domain assumption Proprietary processing by Hove plus follow-up phone interviews produce correct trip boundaries, modes, and purposes.
    Section 2 relies on this hybrid validation, but the algorithms are not public and no inter-rater or accuracy metrics are reported.
  • domain assumption Post-stratification weights calibrated to census marginal distributions make the volunteer sample representative of the Ile-de-France population.
    Section 4.4 describes the calibration; Section 5.1's agreement with INSEE is therefore expected and not an independent check.
  • ad hoc to paper Blurring the first and last points of each trip to H3 resolution 10 centroids is sufficient for GDPR-compliant anonymization while preserving analytical utility.
    Section 3 applies this specific pipeline without a privacy risk analysis; in-trip points remain at full resolution and cell assignment is inconsistent across trips.
  • domain assumption GPS points outside validated trips carry no useful mobility information and can be discarded.
    Section 3 removes all non-trip points; this discards stationary and short intra-location movements, which may matter for some analyses.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The NetMob25 Dataset: A High-resolution Multi-layered View of Individual Mobility in Greater Paris Region." pith.science (2026). https://pith.science/paper/3KBVPX3K

@misc{pith2026250605903,
  author       = {Pith},
  title        = {Pith review of: The NetMob25 Dataset: A High-resolution Multi-layered View of Individual Mobility in Greater Paris Region},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3KBVPX3K}},
  note         = {Machine review of arXiv:2506.05903}
}
read the original abstract

High-quality mobility data remains scarce despite growing interest from researchers and urban stakeholders in understanding individual-level movement patterns. The Netmob25 Data Challenge addresses this gap by releasing a unique GPS-based mobility dataset derived from the EMG 2023 GNSS-based mobility survey conducted in the Ile-de-France region (Greater Paris area), France. This dataset captures detailed daily mobility over a full week for 3,337 volunteer residents aged 16 to 80, collected between October 2022 and May 2023. Each participant was equipped with a dedicated GPS tracking device configured to record location points every 2-3 seconds and was asked to maintain a digital or paper logbook of their trips. All inferred mobility traces were algorithmically processed and validated through follow-up phone interviews. The dataset includes three components: (i) an Individuals database describing demographic, socioeconomic, and household characteristics; (ii) a Trips database with over 80,000 annotated displacements including timestamps, transport modes, and trip purposes; and (iii) a Raw GPS Traces database comprising about 500 million high-frequency points. A statistical weighting mechanism is provided to support population-level estimates. An extensive anonymization pipeline was applied to the GPS traces to ensure GDPR compliance while preserving analytical value. Access to the dataset requires acceptance of the challenge's Terms and Conditions and signing a Non-Disclosure Agreement. This paper describes the survey design, collection protocol, processing methodology, and characteristics of the released dataset.

Figures

Figures reproduced from arXiv: 2506.05903 by the authors.

Figure 1
Figure 1. Structure of the EMG dataset for a single participant ( [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 3
Figure 3. Observed vs. expected department￾level distribution. 75 19.1% 77 11.4% 78 11.6% 91 10.3% 92 13.2% 93 13.1% 94 11.4% 95 9.9% (a) Pie chart of weighted department distribution 75 77 78 91 92 93 94 95 Department 0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 Weighted count 1e6 Observed (weighted) Expected (INSEE) (b) Bar chart of weighted department distribution [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figure 4
Figure 4. Weighted distribution of the population by department in the dataset [PITH_FULL_IMAGE:figures/full_fig_p009_4.png] view at source ↗
Figures from the paper (19 more)
Figure 5
Figure 5. Figure 5: Estimated weighted number of trips per date and per hour. [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]
Figure 6
Figure 6. Figure 6: Number of users and trips per date and per day of the week. [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: Distribution of trip records per individual. [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]
Figure 8
Figure 8. Figure 8: Distribution of trip durations 11 [PITH_FULL_IMAGE:figures/full_fig_p011_8.png]
Figure 9
Figure 9. Figure 9: Distribution of trip durations by transportation mode (capped at 200 min) [PITH_FULL_IMAGE:figures/full_fig_p012_9.png]
Figure 10
Figure 10. Figure 10: Mobility flows between geographical zones in the dataset [PITH_FULL_IMAGE:figures/full_fig_p013_10.png]
Figure 11
Figure 11. Figure 11: Paris IRIS organization, according to INSEE’s original dataset (5264), and the visited [PITH_FULL_IMAGE:figures/full_fig_p014_11.png]
Figure 12
Figure 12. Figure 12: The average number of MoveInside, InComing, and OutGoing trips in each IRIS, per [PITH_FULL_IMAGE:figures/full_fig_p014_12.png]
Figure 13
Figure 13. Figure 13: Number of trips passing by each IRIS per date and per hour. [PITH_FULL_IMAGE:figures/full_fig_p015_13.png]
Figure 14
Figure 14. Figure 14: Spatial distribution of attendance in each IRIS per date and per hour. [PITH_FULL_IMAGE:figures/full_fig_p016_14.png]
Figure 15
Figure 15. Figure 15: Spatial distribution of density (Attendance/IRIS Area) per date and per hour. [PITH_FULL_IMAGE:figures/full_fig_p017_15.png]
Figure 16
Figure 16. Figure 16: Trips’ average length per user. trips are observed on Thursdays, followed by Fridays, Saturdays, and Mondays. Tuesdays exhibit the shortest average trip lengths. In [PITH_FULL_IMAGE:figures/full_fig_p017_16.png]
Figure 17
Figure 17. Figure 17: Empirical cumulative distribution of average trip lengths per individual, across weekdays [PITH_FULL_IMAGE:figures/full_fig_p018_17.png]
Figure 18
Figure 18. Figure 18: Number of unique visited locations (IRIS) per date and per hour. [PITH_FULL_IMAGE:figures/full_fig_p018_18.png]
Figure 19
Figure 19. Figure 19: Number of unique visited IRIS per day period and Per-user distribution. [PITH_FULL_IMAGE:figures/full_fig_p019_19.png]
Figure 20
Figure 20. Figure 20: Weighted number of trips per day of the week, by main transportation mode. [PITH_FULL_IMAGE:figures/full_fig_p020_20.png]
Figure 21
Figure 21. Figure 21: Weighted travel chains starting from four main transportation modes. [PITH_FULL_IMAGE:figures/full_fig_p021_21.png]
Figure 22
Figure 22. Figure 22: Log-scaled spatial distribution of departure locations throughout the day. [PITH_FULL_IMAGE:figures/full_fig_p022_22.png]
Figure 23
Figure 23. Figure 23: Log-scaled spatial distribution of arrival locations throughout the day. [PITH_FULL_IMAGE:figures/full_fig_p023_23.png]

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Paris as a 15-Minute City: An Explainable AI Perspective

    cs.LG 2026-08 unverdicted novelty 6.0 of 10

    Using 70,000 trip segments from the NetMob 2025 Paris dataset, the paper finds that higher local POI availability is associated with less car use and more active travel in central Paris, with weaker effects in outer a...

Reference graph

Works this paper leans on

21 extracted references · 19 canonical work pages · cited by 1 Pith paper

  1. [1]

    A multi-source dataset of urban life in the city of milan and the province of trentino

    Gianni Barlacchi, Marco De Nadai, Roberto Larcher, Antonio Casella, Cristiana Chitic, Giovanni Torrisi, Fabrizio Antonelli, Alessandro Vespignani, Alex Pentland, and Bruno Lepri. A multi-source dataset of urban life in the city of milan and the province of trentino. Scientific Data , 2(1):150055, 2015

    work page 2015

  2. [2]

    Bittencourt and Mariana Giannotti

    Tain´ a A. Bittencourt and Mariana Giannotti. Evaluating the accessibility and availability of public services to reduce inequalities in everyday mobility. Transportation Research Part A: Policy and Practice, 177:103833, 2023

    work page 2023

  3. [3]

    Blondel, Markus Esch, Connie Chan, Fabrice Clerot, Pierre Deville, Etienne Huens, Fr´ ed´ eric Morlot, Zbigniew Smoreda, and Cezary Ziemlicki

    Vincent D. Blondel, Markus Esch, Connie Chan, Fabrice Clerot, Pierre Deville, Etienne Huens, Fr´ ed´ eric Morlot, Zbigniew Smoreda, and Cezary Ziemlicki. Data for development: the d4d challenge on mobile phone data, 2013

    work page 2013

  4. [4]

    Yves-Alexandre de Montjoye, Zbigniew Smoreda, Romain Trinquart, Cezary Ziemlicki, and Vincent D. Blondel. D4d-senegal: The second mobile phone data for development challenge, 2014

    2014

  5. [5]

    The death and life of great italian cities: A mobile phone data perspective

    Marco De Nadai, Jacopo Staiano, Roberto Larcher, Nicu Sebe, Daniele Quercia, and Bruno Lepri. The death and life of great italian cities: A mobile phone data perspective. In Proceedings of the 25th International Conference on World Wide Web , WWW ’16, page 413–423, Republic and Canton of Geneva, CHE, 2016. International World Wide Web Conferences Steering...

    work page 2016

  6. [6]

    Future cities challenge

    Foursquare. Future cities challenge. https://location.foursquare.com/resources/blog/ leadership/how-location-technology-can-drive-urban-innovation/ , 2018. Accessed: 2025-06- 06

    work page 2018

  7. [7]

    Big data for development: preventing the spread of epidemics

    International Telecommunication Union. Big data for development: preventing the spread of epidemics. https://www.itu.int/en/ITU-D/Emergency-Telecommunications/Pages/BigData/ default.aspx, n.d. Accessed: 2025-06-06

    work page 2025

  8. [8]

    El-Yacoubi

    Ghazaleh Khodabandelou, Vincent Gauthier, Marco Fiore, and Mounim A. El-Yacoubi. Estimation of static and dynamic urban populations with mobile network metadata. IEEE Transactions on Mobile Computing, 18(9):2034–2047, 2019

    work page 2019

Show all 21 references
  1. [9]

    Understanding commuting patterns using transit smart card data

    Xiaolei Ma, Congcong Liu, Huimin Wen, Yunpeng Wang, and Yao-Jan Wu. Understanding commuting patterns using transit smart card data. Journal of Transport Geography, 58:135–145, 2017

    2017

  2. [10]

    The netmob23 dataset: A high-resolution multi-region service-level mobile data traffic cartography, 2023

    Orlando E Mart ´ ınez-Durive, Sachit Mishra, Cezary Ziemlicki, Stefania Rubrichi, Zbigniew Smoreda, and Marco Fiore. The netmob23 dataset: A high-resolution multi-region service-level mobile data traffic cartography, 2023. 1https://drive.google.com/file/d/13Jssi358EDWu9Jf4DROT...

    2023

  3. [11]

    Mobility patterns are associated with experienced income segregation in large us cities

    Esteban Moro, Dan Calacci, Xiaowen Dong, and Alex Pentland. Mobility patterns are associated with experienced income segregation in large us cities. Nature Communications, 12(1):4633, 2021

    2021

  4. [12]

    M. M. Nyhan, I. Kloog, R. Britter, C. Ratti, and P. Koutrakis. Quantifying population exposure to air pollution using individual mobility patterns inferred from mobile phone data. Journal of Exposure Science & Environmental Epidemiology , 29(2):238–247, 2019

    2019

  5. [13]

    BT-Q1000XT GPS Travel Recorder User Manual , 2010

    Qstarz International Co. BT-Q1000XT GPS Travel Recorder User Manual , 2010. Accessed: 2025-05- 27

    2010

  6. [14]

    Enquˆ ete r´ egionale sur la mobilit´ e des franciliens

    Institut Paris Region. Enquˆ ete r´ egionale sur la mobilit´ e des franciliens. https://www.institutparisregion.fr/mobilite-et-transports/deplacements/ enquete-regionale-sur-la-mobilite-des-franciliens/ , 2023. Accessed: 2025-06-03

    2023

  7. [15]

    Slides: Enquˆ ete mobilit´ e par gnss (emg) premiers r´ esultats et potentiel des bases de donn´ ees

    Institut Paris Region. Slides: Enquˆ ete mobilit´ e par gnss (emg) premiers r´ esultats et potentiel des bases de donn´ ees. https://infogram.com/premiers-resultats-emg-1hxj48mp3xy3q2v , 2023. Accessed: 2025-06-03

    2023

  8. [16]

    Data for refugees: The d4r challenge on mobility of syrian refugees in turkey, 2018

    Albert Ali Salah, Alex Pentland, Bruno Lepri, Emmanuel Letouze, Patrick Vinck, Yves-Alexandre de Montjoye, Xiaowen Dong, and Ozge Dagdelen. Data for refugees: The d4r challenge on mobility of syrian refugees in turkey, 2018

    2018

  9. [17]

    Estimation of urban zonal speed dynamics from user-activity-dependent positioning data and regional paths

    Manon Seppecher, Ludovic Leclercq, Angelo Furno, Delphine Lejri, and Thamara Vieira da Rocha. Estimation of urban zonal speed dynamics from user-activity-dependent positioning data and regional paths. Transportation Research Part C: Emerging Technologies, 129:103183, 2021

    2021

  10. [18]

    The netmob25 dataset: A high-resolution multi-layered view of individual mobility in greater paris region

    NetMob 2025 Data Challenge Team. The netmob25 dataset: A high-resolution multi-layered view of individual mobility in greater paris region. https://gitlab.inria.fr/netmob2025/data-challenge/ -/blob/main/docs/NetMob25%20Dataset%20Slides.pdf, 2025. Accessed: 2025-06-02

    2025

  11. [19]

    News or social media? socio-economic divide of mobile service consumption

    I˜ naki Ucar, Marco Gramaglia, Marco Fiore, Zbigniew Smoreda, and Esteban Moro. News or social media? socio-economic divide of mobile service consumption. Journal of the Royal Society Interface , 18(185):20210350, December 2021

    2021

  12. [20]

    Aggregated mobility and density data for the netmob 2024 data chal- lenge

    World Bank. Aggregated mobility and density data for the netmob 2024 data chal- lenge. https://datacatalog.worldbank.org/search/dataset/0066094/aggregated_mobility_and_ density_data_for_the_netmob_2024_data_challenge, 2024. Accessed: 2025-06-06

    2024

  13. [21]

    Jones, P

    Takahiro Yabe, Nicholas K.W. Jones, P. Suresh C. Rao, Marta C. Gonzalez, and Satish V. Ukkusuri. Mobile phone location data for disasters: A review from natural hazards and epidemics. Computers, Environment and Urban Systems , 94:101777, 2022. 26

    2022

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.