CoChat
Sign Up
Cited by 105Human Mobility and Location-Based AnalysisComplex Systems and Time Series AnalysisData Mining Algorithms and Applications

The DOI matches its Crossref record; 39 of the 39 references it cites were found in a registry or an index.

Sequences of purchases in credit card data reveal lifestyles in urban populations

Open access
Di Clemente, Riccardo, Luengo-Oroz, Miguel, Travizano, Matias, Xu, Sharon, Vaitla, Bapu, González, Marta C.·2018·Nature Communications
DOI 10.1038/s41467-018-05690-8PMC PMC6102281
Open this paper in CoChat — free accountOpen PDF

Abstract

Zipf-like distributions characterize a wide set of phenomena in physics, biology, economics, and social sciences. In human activities, Zipf's law describes, for example, the frequency of appearance of words in a text or the purchase types in shopping patterns. In the latter, the uneven distribution of transaction types is bound with the temporal sequences of purchases of individual choices. In this work, we define a framework using a text compression technique on the sequences of credit card purchases to detect ubiquitous patterns of collective behavior. Clustering the consumers by their similarity in purchase sequences, we detect five consumer groups. Remarkably, post checking, individuals in each group are also similar in their age, total expenditure, gender, and the diversity of their social and mobility networks extracted from their mobile phone records. By properly deconstructing transaction data with Zipf-like distributions, this method uncovers sets of significant sequences that reveal insights on collective human behavior.

Overview

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

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

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

Integrity signals

Registry

Crossref · 10.1038/s41467-018-05690-8

Cites

39 of 39 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Transaction frequency by type and their demographics. a User median expenditure per month in CCR transactions vs. the average monthly wage in their district of residence. The color and the number represent different districts of Mexico City (see Supplementary Figure 1), and the size of the circles is proportional to the number of users in the district. b Transactions by type as defined by MCC29. c Comparison of frequencies by transaction types (same as in b) separating users in groups according to their gender, income, and age. The share of transaction frequency is distributed similarly among different groups. The icons used in this figure are work of Azaze11o/Shutterstock.com
Fig. 1 Transaction frequency by type and their demographics. a User median expenditure per month in CCR transactions vs. the average monthly wage in their district of residence. The color and the number represent different districts of Mexico City (see Supplementary Figure 1), and the size of the circles is proportional to the number of users in the district. b Transactions by type as defined by MCC29. c Comparison of frequencies by transaction types (same as in b) separating users in groups according to their gender, income, and age. The share of transaction frequency is distributed similarly among different groups. The icons used in this figure are work of Azaze11o/Shutterstock.comCC BY · PubMed Central
Fig. 2 Methods and metrics. a Schematic representation of the Sequitur’s algorithm applied to a sequence of transactions of one user to detect words and identify significant transaction sequences in the data set. b Calculation of the similarity between two users (left) based on the Jaccard index of their significant sequences to define the matrix of users’ similarity (right). Group of users are detected based on similar sequences of transactions. c Lifestyle representation based on sample users 1 and 2 of b. d Example of traces of CDR and CCR data for the user. e Metrics adopted for the analysis of CDR data. The icons used in this figure are work of Azaze11o/Shutterstock.com
Fig. 2 Methods and metrics. a Schematic representation of the Sequitur’s algorithm applied to a sequence of transactions of one user to detect words and identify significant transaction sequences in the data set. b Calculation of the similarity between two users (left) based on the Jaccard index of their significant sequences to define the matrix of users’ similarity (right). Group of users are detected based on similar sequences of transactions. c Lifestyle representation based on sample users 1 and 2 of b. d Example of traces of CDR and CCR data for the user. e Metrics adopted for the analysis of CDR data. The icons used in this figure are work of Azaze11o/Shutterstock.comCC BY · PubMed Central
Fig. 3 Semantic analysis of transaction sequences. a Probability density function plot of the occurrence of words {wi} and its complementary cumulative distribution; the probability distribution words manifest a power-law behavior p(wi)∝xi^(-1.70), with xi frequency of the {wi} and Kolmogorov–Smirnov distance Dn = 0.014. b Distribution of the occurrence of words in the transaction sequences by the word length. c Inter-time transactions between purchases. The purchases within each word are more likely to occur within a day with respect to two random consecutive transactions. c (inside) Moreover, the purchase time to accomplish a word completely is less with respect to two random consecutive transactions. d, e Examples of words composed by two and three codes, respectively, ordered by the number of occurrences. The icons used in this figure are work of Azaze11o/Shutterstock.com
Fig. 3 Semantic analysis of transaction sequences. a Probability density function plot of the occurrence of words {wi} and its complementary cumulative distribution; the probability distribution words manifest a power-law behavior p(wi)∝xi^(-1.70), with xi frequency of the {wi} and Kolmogorov–Smirnov distance Dn = 0.014. b Distribution of the occurrence of words in the transaction sequences by the word length. c Inter-time transactions between purchases. The purchases within each word are more likely to occur within a day with respect to two random consecutive transactions. c (inside) Moreover, the purchase time to accomplish a word completely is less with respect to two random consecutive transactions. d, e Examples of words composed by two and three codes, respectively, ordered by the number of occurrences. The icons used in this figure are work of Azaze11o/Shutterstock.comCC BY · PubMed Central
Fig. 4 Identified lifestyles I. a Groups based on their spending habits. We show the top 10 most frequent spending sequences of the users in each group, representing more than 30% of users’ shopping routines. The percentage of the total users in each group is shown in the bottom-right corner. b Comparison of the median of socio-demographic variables within each group with respect to the median of all users is in red. (The color of the radar plot identifies the spending habits in a.) c Distribution of individual characteristics among users: gender, radius of gyration, mobility diversity, social diversity, median expenditure by month, average distance traveled from the center of residence zipcode to the city center, and age (see Supplementary Figures 11–16, 21 for further information). The icons used in this figure are work of Azaze11o/Shutterstock.com
Fig. 4 Identified lifestyles I. a Groups based on their spending habits. We show the top 10 most frequent spending sequences of the users in each group, representing more than 30% of users’ shopping routines. The percentage of the total users in each group is shown in the bottom-right corner. b Comparison of the median of socio-demographic variables within each group with respect to the median of all users is in red. (The color of the radar plot identifies the spending habits in a.) c Distribution of individual characteristics among users: gender, radius of gyration, mobility diversity, social diversity, median expenditure by month, average distance traveled from the center of residence zipcode to the city center, and age (see Supplementary Figures 11–16, 21 for further information). The icons used in this figure are work of Azaze11o/Shutterstock.comCC BY · PubMed Central
Fig. 5 Identified lifestyles II. a Cell towers residual activity by clusters. b Clusters’ homophily. As expected, each user tends to contact the users that belong to the same clusters or cluster 5 “uncategorized,” which is the cluster with the highest number of users. Remarkably, there is a slight preference to contact cluster 2, the homemakers, which represent the oldest group. c Distribution of returners and explorers across the clusters (see Supplementary Figure 11–16, 21 for further information). Maps in this figure were created using the software QGIS using OpenStreetMap data
Fig. 5 Identified lifestyles II. a Cell towers residual activity by clusters. b Clusters’ homophily. As expected, each user tends to contact the users that belong to the same clusters or cluster 5 “uncategorized,” which is the cluster with the highest number of users. Remarkably, there is a slight preference to contact cluster 2, the homemakers, which represent the oldest group. c Distribution of returners and explorers across the clusters (see Supplementary Figure 11–16, 21 for further information). Maps in this figure were created using the software QGIS using OpenStreetMap dataCC BY · PubMed Central

Verification

Claims the paper makes

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

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

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 39 references

Research works printed but not linked: 3

  1. Solomon, M. R., Dahl, D. W., White, K., Zaichkowsky, J. L. & Polegato, R. Consumer Behavior: Buying, Having, and Being, Vol. 10 (Pearson, Upper Saddle River, 2014).
  2. Yoshimura, Y., Sobolevsky, S., Bautista Hobin, J. N., Ratti, C. & Blat, J. Urban association rules: uncovering linked trips for shopping behavior. Environ. Plan. B 45, 367–385 (2016).
  3. Blei, D. M., Ng, A. Y. & Jordan, M. I. Latent dirichlet allocation. J. Mach. Learn. Res. 3, 993–1022 (2003).

Cited by

Sorted by citations

See all 105

Related

Metadata

Authors
Di Clemente, R., Luengo-Oroz, M., Travizano, M., Xu, S., Vaitla, B., & González, M. C.
Affiliations
  • United Nations (US)
  • Massachusetts Institute of Technology (US)
  • Harvard University (US)
  • Lawrence Berkeley National Laboratory (US)
Corresponding author
González, Marta C.
ORCID
Venue
Nature Communications
Publisher
Springer Science and Business Media LLC
Published
2018-08-20
Volume
9
Issue
1
Pages
3330-3330
ISSN
2041-1723 (ISSN-L)
DOI
10.1038/s41467-018-05690-8
PMC
PMC6102281
OpenAlex
W2594767424
Cited by
105
References
39
Citations by year
2026: 7 · 2025: 11 · 2024: 12 · 2023: 15 · 2022: 16 · 2021: 20 · 2020: 10 · 2019: 10 · 2018: 4
FWCI
13.61
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)
  • arXiv (Cornell University) (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
  • DSpace@MIT (Massachusetts Institute of Technology) (submittedVersion, cc-by-nc)
  • eScholarship (California Digital Library) (submittedVersion, cc-by)
  • Open Research Exeter (University of Exeter) (submittedVersion, cc-by)
Funders
  • Royal Society (NF170505)
  • Bill and Melinda Gates Foundation (OPP1141325)
  • United Nations Population Fund (UNF-15-738)
Topics
  • Social Sciences › Transportation › Human Mobility and Location-Based Analysis
  • Social Sciences › Economics, Econometrics and Finance › Economics and Econometrics › Complex Systems and Time Series Analysis
  • Physical Sciences › Computer Science › Information Systems › Data Mining Algorithms and Applications
Keywords
credit card transactions, Zipf's law, consumer lifestyle, consumer clustering, collective behavior, text compression, social networks, mobility networks, urban population
MeSH terms
Consumer Behavior, Humans, Life Style, Semantics, Urban Population, 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