pith. sign in

arxiv: 2605.18777 · v1 · pith:MRAVWJ2Qnew · submitted 2026-04-23 · 💻 cs.SI · cs.CV

XFlowMap: Cross-Scale Generalization and Mapping of Massive Origin-Destination Data

Diansheng Guo , Hai Jin This is my paper

Pith reviewed 2026-05-20 23:45 UTC · model grok-4.3

classification 💻 cs.SI cs.CV
keywords origin-destination dataflow mappingcross-scale generalizationscan statisticsmobility visualizationcartographic generalizationspatial pattern detection
0
0 comments X

The pith

XFlowMap detects salient origin-destination flow patterns at their natural scales and renders them with a single integrated symbol.

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

The paper introduces XFlowMap to address cluttered flow maps from massive origin-destination datasets by automatically locating meaningful patterns that appear at different spatial scales. It combines a scan-statistic procedure that identifies and generalizes cross-scale flow clusters without preset units or manual tuning, followed by a new visualization symbol that encodes location, direction, strength, and the relevant scales together. A sympathetic reader would care because existing approaches to mobility mapping often force fixed aggregations that hide or distort multi-scale structures in data such as migration or commuting flows. If the method performs as described, analysts gain the ability to produce readable static and interactive maps that reveal high-level flow structures even in large, sparse, or noisy collections. The framework applies to both area-based and point-based origin-destination inputs and supports comparison across stratified subsets of the data.

Core claim

The framework detects salient flow patterns at their appropriate origin and destination scales, extracts high-level structures through automated generalization, and generates a new flow map representation that integrates location, direction, strength, and OD scales in one symbol. A scan-statistic-based procedure evaluates and generalizes the cross-scale clusters. The resulting maps support holistic interpretation of complex origin-destination structures for both static presentation and interactive exploration.

What carries the argument

Scan-statistic-based procedure for cross-scale flow cluster detection and generalization, together with a novel flow symbol that encodes location, direction, strength, and origin-destination scales in a single representation.

If this is right

  • Produces clear, information-rich flow maps for large mobility datasets.
  • Supports both static presentation and interactive exploration of the detected patterns.
  • Handles area-based and point-based origin-destination data equally.
  • Remains effective on sparse and noisy input collections.
  • Allows direct comparative mapping across stratified subsets of the flow data.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same scale-detection logic could be tested on logistics or internet-traffic flows to surface operational patterns that only emerge at particular geographic resolutions.
  • Pairing the scan-statistic step with automated parameter tuning might reduce sensitivity to the choice of significance thresholds in very large dynamic datasets.
  • The integrated symbol design offers a template for other network visualizations where direction, magnitude, and hierarchical grouping must be shown without separate legends.

Load-bearing premise

The scan-statistic-based procedure accurately detects salient flow patterns at their appropriate origin and destination scales without predefined aggregation units or manual intervention.

What would settle it

Apply the procedure to synthetic origin-destination data containing known multi-scale clusters at specific scales and check whether the output recovers those exact clusters at the correct scales while suppressing noise-induced artifacts.

Figures

Figures reproduced from arXiv: 2605.18777 by Diansheng Guo, Hai Jin.

Figure 1
Figure 1. Figure 1: : Illustration of a candidate flow pattern (cluster), which is defined by a [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: : Cross-scale search of candidate flow clusters. For each observed flow (black [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: The empirical distribution of maxLGLR with 100 permutations of the senior migration data (Section 5) and the fitted Gumbel curves in red (location µ = 16.12 and scale β = 1.06). In practice, statistical testing is optional for the purpose of cartographic general￾ization. Because clusters are ranked by LGLR, the top clusters selected for mapping typically represent only a small subset of all statistically s… view at source ↗
Figure 4
Figure 4. Figure 4: : Flow symbol design for representing a cross-scale flow cluster. [PITH_FULL_IMAGE:figures/full_fig_p012_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: : Experiments with synthetic datasets. (A) 600 flows in eight clusters; (B) [PITH_FULL_IMAGE:figures/full_fig_p014_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: : Non-overlapping significant flow clusters of senior migrants (age 65-69) with [PITH_FULL_IMAGE:figures/full_fig_p016_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: : Six selected clusters from Figure 6, with origin/destination circles shown, [PITH_FULL_IMAGE:figures/full_fig_p016_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: : A zoomed-in flow map with additional details ( [PITH_FULL_IMAGE:figures/full_fig_p017_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: : Flow clusters of young migrants (age 25-29) with [PITH_FULL_IMAGE:figures/full_fig_p018_9.png] view at source ↗
read the original abstract

Mapping large origin-destination (OD) datasets remains challenging because flow maps become cluttered, meaningful patterns occur at multiple spatial scales, and existing flow-mapping approaches frequently rely on predefined aggregation units or manual generalization. This paper presents XFlowMap, a framework for the cross-scale generalization and mapping of massive OD data. Specifically, the framework integrates cross-scale flow pattern (cluster) detection, automated flow map generalization, and a new cartographic representation for analyzing and visualizing complex origin-destination flow structures. The approach detects salient flow patterns at their appropriate origin and destination scales, extracts high-level structures, and generates a new flow map representation that supports holistic interpretation of complex origin-destination flow patterns. A scan-statistic-based procedure is developed to evaluate and generalize cross-scale flow clusters. The detected clusters are then visualized using a novel flow symbol that integrates location, direction, strength, and OD scales in a single representation. The framework supports both area-based and point-based OD data, is robust to sparse and noisy datasets, and enables comparative mapping of stratified flow data. Experiments with synthetic data and U.S. migration data demonstrate that the method effectively extracts meaningful cross-scale flow patterns and produces clear, information-rich flow maps for large mobility datasets, supporting both static presentation and interactive exploration.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper introduces XFlowMap, a framework for cross-scale generalization and mapping of massive origin-destination (OD) data. It integrates a scan-statistic-based procedure for detecting salient flow clusters at appropriate origin and destination scales without predefined aggregation units or manual intervention, automated generalization of high-level structures, and a novel flow symbol that encodes location, direction, strength, and OD scales in one representation. The approach is claimed to handle area- and point-based data, be robust to sparse/noisy inputs, and support comparative stratified mapping; experiments on synthetic data and U.S. migration data are presented as demonstrating effective extraction of meaningful cross-scale patterns and production of clear, information-rich maps for static and interactive use.

Significance. If the central claims on accurate scale detection hold, the work would advance spatial visualization and mobility analytics by reducing reliance on manual or fixed-scale aggregation in flow mapping, with potential utility for large-scale transportation, migration, and urban studies where multi-scale patterns are common.

major comments (2)
  1. [Abstract] Abstract: the central claim that the scan-statistic procedure 'detects salient flow patterns at their appropriate origin and destination scales' without predefined units is load-bearing yet unsupported; no quantitative metrics (e.g., precision/recall on recovered scales against ground truth, sensitivity analysis on window sizes or null-model assumptions) are supplied to confirm scales are meaningful rather than artifacts of the scanning process.
  2. [Experiments] Experiments section: while synthetic and U.S. migration datasets are invoked to show effectiveness, the description provides no error analysis, baseline comparisons (e.g., to hierarchical clustering or fixed-scale methods), or validation that detected scales match appropriate ground-truth scales, leaving the robustness claims without direct empirical grounding.
minor comments (2)
  1. [Methods] Clarify the exact definition and parameterization of the scan-statistic (e.g., window sizes, null model) in the methods section to allow reproducibility.
  2. [Visualization] The novel flow symbol description would benefit from an explicit legend or encoding table showing how OD scales are visually distinguished from strength and direction.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for the constructive feedback on our manuscript describing XFlowMap. The comments highlight important aspects of empirical validation for the scale-detection claims. We address each major comment below with additional context from the work and indicate planned revisions to strengthen the presentation.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim that the scan-statistic procedure 'detects salient flow patterns at their appropriate origin and destination scales' without predefined units is load-bearing yet unsupported; no quantitative metrics (e.g., precision/recall on recovered scales against ground truth, sensitivity analysis on window sizes or null-model assumptions) are supplied to confirm scales are meaningful rather than artifacts of the scanning process.

    Authors: The scan-statistic procedure identifies significant flow clusters by evaluating a range of candidate scales and locations using a null model of randomized OD flows, without requiring predefined aggregation units. Synthetic experiments inject known multi-scale patterns and recover them, providing implicit support. We agree, however, that explicit quantitative metrics would better substantiate the claim. In revision we will add precision/recall for scale recovery on synthetic data, plus sensitivity analysis on window sizes and null-model parameters, and will adjust the abstract wording for precision if needed. revision: yes

  2. Referee: [Experiments] Experiments section: while synthetic and U.S. migration datasets are invoked to show effectiveness, the description provides no error analysis, baseline comparisons (e.g., to hierarchical clustering or fixed-scale methods), or validation that detected scales match appropriate ground-truth scales, leaving the robustness claims without direct empirical grounding.

    Authors: Synthetic experiments use data with injected ground-truth clusters at known scales to demonstrate recovery, while U.S. migration results are assessed qualitatively for interpretability. We acknowledge the value of explicit baselines and error metrics. Revision will incorporate comparisons against fixed-scale aggregation and hierarchical clustering, plus quantitative error analysis on the synthetic case. For real-world data, ground-truth scales are unavailable by nature, so we will expand robustness tests to sparsity and noise instead. revision: partial

standing simulated objections not resolved
  • Quantitative scale-recovery metrics against ground truth cannot be computed for the U.S. migration dataset because no independent ground-truth scale labels exist for real mobility data.

Circularity Check

0 steps flagged

No circularity: new scan-statistic procedure and flow symbols introduced without self-referential reduction.

full rationale

The paper describes a framework that integrates a scan-statistic-based procedure for detecting cross-scale OD clusters, automated generalization, and a novel flow symbol. No equations, fitted parameters, or derivations are presented that reduce the extracted patterns or visualizations to inputs by construction. The method is applied to synthetic data and U.S. migration data for validation, with claims resting on the independent performance of the new procedure rather than any self-definition, self-citation chain, or renaming of prior results. This constitutes a self-contained methodological contribution.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The framework relies on standard assumptions of scan statistics for cluster detection and the utility of a new visual encoding; no explicit free parameters or invented physical entities are named in the abstract.

axioms (1)
  • domain assumption Scan statistics can identify statistically significant spatial clusters in flow data at multiple scales without predefined units.
    Invoked in the description of the cross-scale flow pattern detection procedure.
invented entities (1)
  • Novel flow symbol integrating location, direction, strength, and OD scales no independent evidence
    purpose: To represent complex origin-destination structures in a single graphic element
    Introduced as a new cartographic representation; no independent evidence outside the framework is provided in the abstract.

pith-pipeline@v0.9.0 · 5754 in / 1370 out tokens · 37289 ms · 2026-05-20T23:45:16.888749+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

52 extracted references · 52 canonical work pages

  1. [1]

    Revealing

    Andrienko, Gennady and Andrienko, Natalia and Fuchs, Georg and Wood, Jo , year = 2017, month = sep, journal =. Revealing. doi:10.1109/TVCG.2016.2616404 , urldate =

    work page doi:10.1109/tvcg.2016.2616404 2017

  2. [2]

    Andrienko, Gennady and Andrienko, Natalia and Chen, Wei and Maciejewski, Ross and Zhao, Ye , year = 2017, month = aug, journal =. Visual. doi:10.1109/TITS.2017.2683539 , urldate =

    work page doi:10.1109/tits.2017.2683539 2017

  3. [3]

    Buchin, Kevin and Speckmann, Bettina and Verbeek, Kevin , year = 2011, month = dec, journal =. Flow. doi:10.1109/TVCG.2011.202 , urldate =

    work page doi:10.1109/tvcg.2011.202 2011

  4. [4]

    Geometry-

    Cui, Weiwei and Zhou, Hong and Qu, Huamin and Wong, Pak Chung and Li, Xiaoming , year = 2008, month = nov, journal =. Geometry-. doi:10.1109/TVCG.2008.135 , urldate =

    work page doi:10.1109/tvcg.2008.135 2008

  5. [5]

    Evaluation of

    Duczmal, Luiz and Kulldorff, Martin and Huang, Lan , year = 2006, month = jun, journal =. Evaluation of. doi:10.1198/106186006X112396 , urldate =

    work page doi:10.1198/106186006x112396 2006

  6. [6]

    Eubank, Stephen and Guclu, Hasan and Anil Kumar, V. S. and Marathe, Madhav V. and Srinivasan, Aravind and Toroczkai, Zolt. Modelling. Nature , volume =. doi:10.1038/nature02541 , urldate =

    work page doi:10.1038/nature02541

  7. [7]

    and Hastings, N

    Evans, M. and Hastings, N. and Peacock, B. , year = 2000, publisher =. Statistical

    work page 2000

  8. [8]

    Overlaying

    Fekete, Jean-Daniel and Wang, David and Dang, Niem and Plaisant, Catherine , editor =. Overlaying. Information

    work page

  9. [9]

    Discovering

    Gao, Song and Liu, Yu and Wang, Yaoli and Ma, Xiujun , year = 2013, journal =. Discovering. doi:10.1111/tgis.12042 , abstract =

    work page doi:10.1111/tgis.12042 2013

  10. [10]

    , year = 1958, publisher =

    Gumbel, E.J. , year = 1958, publisher =. Statistics of

    work page 1958

  11. [11]

    Detecting

    Guo, Diansheng and Chen, Chao , year = 2014, journal =. Detecting. doi:10.1111/tgis.12101 , urldate =

    work page doi:10.1111/tgis.12101 2014

  12. [12]

    Discovering

    Guo, Diansheng and Zhu, Xi and Jin, Hai and Gao, Peng and Andris, Clio , year = 2012, month = jun, journal =. Discovering. doi:10.1111/j.1467-9671.2012.01344.x , urldate =

    work page doi:10.1111/j.1467-9671.2012.01344.x 2012

  13. [13]

    Guo, Diansheng , year = 2009, journal =. Flow

    work page 2009

  14. [14]

    Guo, Diansheng and Zhu, Xi , year = 2014, journal =. Origin-. doi:10.1109/TVCG.2014.2346271 , abstract =

    work page doi:10.1109/tvcg.2014.2346271 2014

  15. [15]

    Guo, D , year = 2007, journal =. Visual. doi:10.1080/13658810701349037 , abstract =

    work page doi:10.1080/13658810701349037 2007

  16. [16]

    Holten, Danny and Isenberg, Petra and. An. 2011. doi:10.1109/PACIFICVIS.2011.5742390 , urldate =

    work page doi:10.1109/pacificvis.2011.5742390 2011

  17. [17]

    Visualization of Multi-Variate Scientific Data

    Holten, Danny and Van Wijk, Jarke J. , year = 2009, journal =. Force-. doi:10.1111/j.1467-8659.2009.01450.x , urldate =

    work page doi:10.1111/j.1467-8659.2009.01450.x 2009

  18. [18]

    Ripley's

    Kan, Zihan and Kwan, Mei-Po and Tang, Luliang , year = 2022, journal =. Ripley's. doi:10.1111/gean.12300 , urldate =

    work page doi:10.1111/gean.12300 2022

  19. [19]

    Smoothing

    Koylu, Caglar and Guo, Diansheng , year = 2013, month = sep, journal =. Smoothing. doi:10.1016/j.compenvurbsys.2013.03.001 , urldate =

    work page doi:10.1016/j.compenvurbsys.2013.03.001 2013

  20. [20]

    Kulldorff, Martin and Huang, Lan and Pickle, Linda and Duczmal, Luiz , year = 2006, journal =. An. doi:10.1002/sim.2490 , urldate =

    work page doi:10.1002/sim.2490 2006

  21. [21]

    Prospective

    Kulldorff, Martin , year = 2001, journal =. Prospective. doi:10.1111/1467-985X.00186 , urldate =

    work page doi:10.1111/1467-985x.00186 2001

  22. [22]

    Kulldorff, Martin and Heffernan, Richard and Hartman, Jessica and Assun. A. PLOS Medicine , volume =. doi:10.1371/journal.pmed.0020059 , urldate =

    work page doi:10.1371/journal.pmed.0020059

  23. [23]

    doi:10.1080/13658816.2021.1899184 , urldate =

    Liu, Qiliang and Yang, Jie and Deng, Min and Song, Ci and Liu, Wenkai , year = 2022, month = feb, journal =. doi:10.1080/13658816.2021.1899184 , urldate =

    work page doi:10.1080/13658816.2021.1899184 2022

  24. [24]

    Uncovering

    Liu, Yu and Sui, Zhengwei and Kang, Chaogui and Gao, Yong , year = 2014, month = jan, journal =. Uncovering. doi:10.1371/journal.pone.0086026 , urldate =

    work page doi:10.1371/journal.pone.0086026 2014

  25. [25]

    Understanding

    Liu, Yu and Kang, Chaogui and Gao, Song and Xiao, Yu and Tian, Yuan , year = 2012, month = oct, journal =. Understanding. doi:10.1007/s10109-012-0166-z , urldate =

    work page doi:10.1007/s10109-012-0166-z 2012

  26. [26]

    Identifying

    Mitchell, William and Watts, Martin , year = 2010, journal =. Identifying. doi:10.1111/j.1745-5871.2009.00631.x , urldate =

    work page doi:10.1111/j.1745-5871.2009.00631.x 2010

  27. [27]

    , year = 1984, journal =

    Openshaw, S. , year = 1984, journal =. The

    work page 1984

  28. [28]

    Phan, Doantam and Xiao, Ling and Yeh, Ron and Hanrahan, Pat and Winograd, Terry , year = 2005, month = oct, pages =. Flow. doi:10.1109/INFVIS.2005.1532150 , urldate =

    work page doi:10.1109/infvis.2005.1532150 2005

  29. [29]

    , year = 1997, month = oct, journal =

    Poon, Jessie P. , year = 1997, month = oct, journal =. The. doi:10.1111/j.1944-8287.1997.tb00096.x , urldate =

    work page doi:10.1111/j.1944-8287.1997.tb00096.x 1997

  30. [30]

    Rae, Alasdair , year = 2009, month = may, journal =. From. doi:10.1016/j.compenvurbsys.2009.01.007 , urldate =

    work page doi:10.1016/j.compenvurbsys.2009.01.007 2009

  31. [31]

    and Rice, John A

    Rice, John A. and Rice, John A. , year = 2007, volume =. Mathematical

    work page 2007

  32. [32]

    , year = 2011, journal =

    Scheepens, Roeland and Willems, Niels and Van De Wetering, Huub and Andrienko, Gennady and Andrienko, Natalia and Van Wijk, Jarke J. , year = 2011, journal =. Composite. doi:10.1109/TVCG.2011.181 , abstract =

    work page doi:10.1109/tvcg.2011.181 2011

  33. [33]

    Interactive

    Scheepens, Roeland and Willems, Niels and. Interactive. IEEE Computer Graphics and Applications , volume =. doi:10.1109/MCG.2011.88 , urldate =

    work page doi:10.1109/mcg.2011.88 2011

  34. [34]

    International Journal of Geographical Information Science , volume =

    L-Function of Geographical Flows , author =. International Journal of Geographical Information Science , volume =. doi:10.1080/13658816.2020.1749277 , urldate =

    work page doi:10.1080/13658816.2020.1749277 2020

  35. [35]

    Zhang, Y

    Detecting Arbitrarily Shaped Clusters in Origin-Destination Flows Using Ant Colony Optimization , author =. International Journal of Geographical Information Science , volume =. doi:10.1080/13658816.2018.1516287 , urldate =

    work page doi:10.1080/13658816.2018.1516287 2018

  36. [36]

    Tao, Ran and Thill, Jean-Claude , year = 2019, month = oct, journal =. Flow. doi:10.1080/13658816.2019.1608362 , urldate =

    work page doi:10.1080/13658816.2019.1608362 2019

  37. [37]

    Proceedings of the 3rd

    Tao, Ran and Thill, Jean-Claude and Depken, Craig and Kashiha, Mona , year = 2017, month = nov, series =. Proceedings of the 3rd. doi:10.1145/3152178.3152189 , urldate =

    work page doi:10.1145/3152178.3152189 2017

  38. [38]

    , year = 1987, month = jan, journal =

    Tobler, Waldo R. , year = 1987, month = jan, journal =. Experiments. doi:10.1559/152304087783875273 , urldate =

    work page doi:10.1559/152304087783875273 1987

  39. [39]

    Tobler, W. R. , year = 1981, journal =. A. doi:10.1111/j.1538-4632.1981.tb00711.x , urldate =

    work page doi:10.1111/j.1538-4632.1981.tb00711.x 1981

  40. [40]

    Tobler, Waldo , year = 1976, month = jan, journal =. Spatial. doi:10.2190/VAKC-3GRF-3XUG-WY4W , urldate =

    work page doi:10.2190/vakc-3grf-3xug-wy4w 1976

  41. [41]

    Ware, Colin and Kelley, John G. W. and Pilar, David , year = 2014, month = oct, journal =. Improving the. doi:10.1175/BAMS-D-13-00135.1 , urldate =

    work page doi:10.1175/bams-d-13-00135.1 2014

  42. [42]

    Visualisation of

    Wood, Jo and Dykes, Jason and Slingsby, Aidan , year = 2010, month = may, journal =. Visualisation of. doi:10.1179/000870410X12658023467367 , urldate =

    work page doi:10.1179/000870410x12658023467367 2010

  43. [43]

    Spatiotemporal

    Yan, Xiaorui and Pei, Tao and Shu, Hua and Song, Ci and Wu, Mingbo and Fang, Zidong and Chen, Jie , year = 2023, month = jul, journal =. Spatiotemporal. doi:10.1080/13658816.2023.2204345 , urldate =

    work page doi:10.1080/13658816.2023.2204345 2023

  44. [44]

    Yao, Xin and Zhu, Di and Gao, Yong and Wu, Lun and Zhang, Pengcheng and Liu, Yu , year = 2018, journal =. A. doi:10.1109/ACCESS.2018.2864662 , urldate =

    work page doi:10.1109/access.2018.2864662 2018

  45. [45]

    , year = 2002, journal =

    Young, David A. , year = 2002, journal =. A. doi:10.1525/aa.2002.104.1.138 , urldate =

    work page doi:10.1525/aa.2002.104.1.138 2002

  46. [46]

    and Shen, Q

    Zeng, W. and Shen, Q. and Jiang, Y. and Telea, A. , year = 2019, journal =. Route-. doi:10.1111/cgf.13712 , urldate =

    work page doi:10.1111/cgf.13712 2019

  47. [47]

    Zhou, Zhiguang and Meng, Linhao and Tang, Cheng and Zhao, Ying and Guo, Zhiyong and Hu, Miaoxin and Chen, Wei , year = 2019, month = jan, journal =. Visual. doi:10.1109/TVCG.2018.2864503 , urldate =

    work page doi:10.1109/tvcg.2018.2864503 2019

  48. [48]

    Zhu, Xi and Guo, Diansheng , year = 2014, journal =. Mapping. doi:10.1111/tgis.12100 , langid =

    work page doi:10.1111/tgis.12100 2014

  49. [49]

    2014 , school=

    Pattern extraction from spatial data-Statistical and modeling approaches , author=. 2014 , school=

    work page 2014

  50. [50]

    Cartography and Geographic Information Science , volume=

    Design principles for origin-destination flow maps , author=. Cartography and Geographic Information Science , volume=. 2018 , publisher=

    work page 2018

  51. [51]

    IEEE transactions on visualization and computer graphics , volume=

    A visualization system for space-time and multivariate patterns (vis-stamp) , author=. IEEE transactions on visualization and computer graphics , volume=. 2006 , publisher=

    work page 2006

  52. [52]

    Information Visualization , volume=

    Design and evaluation of line symbolizations for origin--destination flow maps , author=. Information Visualization , volume=. 2017 , publisher=

    work page 2017