pith. sign in

arxiv: 1907.11498 · v2 · pith:VLYBKH3Lnew · submitted 2019-07-26 · 💻 cs.HC · cs.CY

Personality is Revealed During Weekends: Towards Data Minimisation for Smartphone Based Personality Classification

Mohammed Khwaja , Aleksandar Matic This is my paper

Pith reviewed 2026-05-24 15:25 UTC · model grok-4.3

classification 💻 cs.HC cs.CY
keywords personality classificationsmartphone datadata minimisationweekend databig five traitsprivacybehavioral modeling
0
0 comments X

The pith

Smartphone data from one or two weekends classifies the five personality traits at 66-71 percent accuracy.

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

The paper tests whether personality classification from smartphone sensors requires weeks or months of data or can succeed with far less. It shows that restricting collection to weekends alone reaches the same accuracy levels reported for longer collection windows. This result directly supports the principle of data minimisation by demonstrating that far smaller volumes of personal behavioral data are enough for the task. A sympathetic reader would care because shorter collection windows could lower privacy risks and increase user willingness to allow such modeling.

Core claim

Data from only one or two weekends is sufficient to achieve state-of-the-art accuracy between 66 percent and 71 percent when classifying the five personality traits using smartphone sensors.

What carries the argument

Weekend-only subsets of smartphone sensor and usage logs used as input features for standard machine-learning classifiers of the Big Five traits.

If this is right

  • Personality classification models can be trained with collection windows reduced from weeks or months to one or two weekends.
  • Smartphone-based personality services can comply with data-minimisation rules while retaining reported accuracy levels.
  • User burden and privacy exposure decrease when data collection is limited to weekends.

Where Pith is reading between the lines

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

  • The finding implies that personality-related behaviors may be more stable or detectable on weekends than on weekdays.
  • Similar weekend-only sampling might be tested for other inference tasks that currently rely on multi-week smartphone data.
  • Designers could explore whether specific weekend activities drive the predictive power rather than overall usage volume.

Load-bearing premise

Weekend smartphone data alone contains enough distinctive behavioral signals to classify personality traits accurately without weekday data or longer collection periods.

What would settle it

A study on a new user cohort that measures classification accuracy below 66 percent when models are trained on only one or two weekends of data would falsify the central result.

read the original abstract

Previous literature has explored automatic personality modelling using smartphone data for its potential to personalise mobile services. Although passive modelling of personality removes the burden of completing lengthy questionnaires, the fact that such models typically require a few weeks or months of personal data can negatively impact user's engagement. In this study, we explore the feasibility of reducing the duration of data collection in the context of personality classification. We found that only one or two weekends can suffice for achieving state-of-the-art accuracy between 66% and 71% for classifying the five personality traits. These results provide lessons for practicing "data minimisation" - a key principle of privacy laws.

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 / 1 minor

Summary. The manuscript claims that smartphone sensor data collected over only one or two weekends is sufficient to reach state-of-the-art classification accuracy (66–71 %) for the five Big-Five personality traits, thereby demonstrating that extended multi-week collection periods are unnecessary and supporting the data-minimisation principle.

Significance. If the reported accuracies are robust, the result would be practically significant for privacy-aware mobile systems: shorter collection windows reduce user burden and regulatory risk while preserving predictive utility. The work supplies an empirical demonstration that weekend behavioural traces alone carry substantial personality signal, which could inform minimal-data designs in human–computer interaction.

major comments (2)
  1. [Abstract] Abstract: the central claim that “only one or two weekends can suffice for achieving state-of-the-art accuracy between 66 % and 71 %” is presented without any accompanying information on sample size (N), number of participants, cross-validation procedure, baseline models, or statistical significance. These details are load-bearing for assessing whether the weekend-only restriction genuinely supports the sufficiency claim.
  2. [Results] Results section (inferred from abstract): the manuscript does not report a direct comparison between weekend-only models and equivalent-length weekday-only or random-day subsets. Without such a control it remains unclear whether the reported performance is attributable to the weekend restriction or simply to any short observation window.
minor comments (1)
  1. Clarify the machine-learning pipeline (feature extraction, classifier, hyper-parameter tuning) and state whether the same pipeline was used for both weekend and longer-period baselines.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on our manuscript. The comments highlight opportunities to strengthen the presentation of key methodological details and the experimental design. We respond to each major comment below.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim that “only one or two weekends can suffice for achieving state-of-the-art accuracy between 66 % and 71 %” is presented without any accompanying information on sample size (N), number of participants, cross-validation procedure, baseline models, or statistical significance. These details are load-bearing for assessing whether the weekend-only restriction genuinely supports the sufficiency claim.

    Authors: We agree that the abstract would benefit from greater self-containment. The full manuscript reports the sample size, participant details, cross-validation procedure, baseline comparisons, and evaluation metrics in the Methods and Results sections. To address this point directly, we will revise the abstract to include a concise statement on the number of participants and the validation approach used. revision: yes

  2. Referee: [Results] Results section (inferred from abstract): the manuscript does not report a direct comparison between weekend-only models and equivalent-length weekday-only or random-day subsets. Without such a control it remains unclear whether the reported performance is attributable to the weekend restriction or simply to any short observation window.

    Authors: The manuscript's primary contribution is to demonstrate that one or two weekends of data suffice to reach the reported accuracy levels, comparable to those achieved with multi-week collection periods in prior work. This directly supports the data-minimisation claim without requiring a claim that weekends are uniquely effective among all short windows. The existing comparison is between weekend-only data and full multi-week traces. A direct weekday-only or random-day control of matched length is not reported and would constitute an extension rather than a necessary control for the stated sufficiency result. revision: no

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper reports an empirical result from smartphone data analysis showing that one or two weekends suffice for 66-71% personality classification accuracy. No equations, derivations, or parameter-fitting steps are described that would reduce any 'prediction' to its own inputs by construction. The central claim is presented as a data-driven observation rather than a self-referential or self-citation-dependent derivation. No load-bearing self-citations, ansatzes, or uniqueness theorems appear in the provided text. This is a standard empirical study whose argument chain is self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Only the abstract is available; no explicit free parameters, axioms, or invented entities are identifiable from the provided text. The work appears to rely on standard machine-learning classification assumptions not detailed here.

pith-pipeline@v0.9.0 · 5632 in / 924 out tokens · 23245 ms · 2026-05-24T15:25:02.925708+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

30 extracted references · 30 canonical work pages

  1. [1]

    In: International conference on user modeling, ada ptation, and personal- ization

    Abel, F., Henze, N., Herder, E., Krause, D.: Interweaving public user profiles on the web. In: International conference on user modeling, ada ptation, and personal- ization. pp. 16–27. Springer (2010)

    work page 2010

  2. [2]

    In: Proceed- ings of the 2015 ACM international joint conference on perva sive and ubiquitous computing

    Canzian, L., Musolesi, M.: Trajectories of depression: u nobtrusive monitoring of depressive states by means of smartphone mobility traces an alysis. In: Proceed- ings of the 2015 ACM international joint conference on perva sive and ubiquitous computing. pp. 1293–1304. ACM (2015)

    work page 2015

  3. [3]

    Jour- nal of Software 12(11), 882–892 (2017)

    Catal, C., et al.: Cross-cultural personality predictio n based on twitter data. Jour- nal of Software 12(11), 882–892 (2017)

    work page 2017

  4. [4]

    I n: Wearable Computers (ISWC), 2011 15th Annual International Symposium on

    Chittaranjan, G., Blom, J., Gatica-Perez, D.: Who’s who w ith big-five: Analyz- ing and classifying personality traits with smartphones. I n: Wearable Computers (ISWC), 2011 15th Annual International Symposium on. pp. 29 –36. IEEE (2011)

    work page 2011

  5. [5]

    Personal and Ubiquitous Computin g 17(3), 433–450 (2013)

    Chittaranjan, G., Blom, J., Gatica-Perez, D.: Mining lar ge-scale smartphone data for personality studies. Personal and Ubiquitous Computin g 17(3), 433–450 (2013)

    work page 2013

  6. [6]

    Cambridge University Press Cambridge, UK: (2009)

    Corr, P.J., Matthews, G.: The Cambridge handbook of perso nality psychology. Cambridge University Press Cambridge, UK: (2009)

    work page 2009

  7. [7]

    The Cambridge hand book of personality psychology pp

    DeYoung, C.G., Gray, J.R.: Personality neuroscience: Ex plaining individual differ- ences in affect, behaviour and cognition. The Cambridge hand book of personality psychology pp. 323–346 (2009)

    work page 2009

  8. [8]

    In: Joint European Conference on Machine Learning and Knowledge Dis- covery in Databases

    Ferwerda, B., Schedl, M.: Personality-based user modeli ng for music recommender systems. In: Joint European Conference on Machine Learning and Knowledge Dis- covery in Databases. pp. 254–257. Springer (2016)

    work page 2016

  9. [9]

    In: The 23rd Intern ational on Intelligent User Interfaces (2018)

    Ferwerda, B., Tkalcic, M.: You are what you post: What the c ontent of instagram pictures tells about users personality. In: The 23rd Intern ational on Intelligent User Interfaces (2018)

    work page 2018

  10. [10]

    Journal of Research i n personality 40(1), 84– 96 (2006)

    Goldberg, L.R., et al.: The international personality i tem pool and the future of public-domain personality measures. Journal of Research i n personality 40(1), 84– 96 (2006)

    work page 2006

  11. [11]

    Personality and Individual Differences 39(2), 317–329 (2005)

    Gow, A.J., Whiteman, M.C., Pattie, A., Deary, I.J.: Gold berg’s ’ipip’ big-five factor markers: Internal consistency and concurrent validation i n scotland. Personality and Individual Differences 39(2), 317–329 (2005)

    work page 2005

  12. [12]

    In: Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems

    Jia, Y., et al.: Personality-targeted gamification: a su rvey study on personality traits and motivational affordances. In: Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems. pp. 2001–2013. ACM (2 016)

    work page 2016

  13. [13]

    In: Proceedings of the 2017 ACM Workshop on Theory-Informed User Modeling for T ailoring and Personalizing Interfaces

    Lee, M.J., Ferwerda, B.: Personalizing online educatio nal tools. In: Proceedings of the 2017 ACM Workshop on Theory-Informed User Modeling for T ailoring and Personalizing Interfaces. pp. 27–30. ACM (2017)

    work page 2017

  14. [14]

    In: Proceedings o f the ACM 2011 con- ference on Computer supported cooperative work

    Minamikawa, A., Yokoyama, H.: Blog tells what kind of per sonality you have: egogram estimation from japanese weblog. In: Proceedings o f the ACM 2011 con- ference on Computer supported cooperative work. pp. 217–22 0. ACM (2011)

    work page 2011

  15. [15]

    Journal of Research in Personali ty 74, 16–22 (2018)

    Mønsted, B., Mollgaard, A., Mathiesen, J.: Phone-based metric as a predictor for basic personality traits. Journal of Research in Personali ty 74, 16–22 (2018)

    work page 2018

  16. [16]

    In: International conference on social computing , behavioral-cultural mod- eling, and prediction

    de Montjoye, Y.A., et al.: Predicting personality using novel mobile phone-based metrics. In: International conference on social computing , behavioral-cultural mod- eling, and prediction. pp. 48–55. Springer (2013)

    work page 2013

  17. [17]

    ACM Transactions on Computer-H uman Interaction (TOCHI) 20(2), 10 (2013) 10 M

    de Oliveira, R., Cherubini, M., Oliver, N.: Influence of p ersonality on satisfaction with mobile phone services. ACM Transactions on Computer-H uman Interaction (TOCHI) 20(2), 10 (2013) 10 M. Khwaja & A. Matic

    work page 2013

  18. [18]

    In: CHI’11 Extended Abstracts on Human Factors in Com puting Systems

    de Oliveira, R., et al.: Towards a psychographic user mod el from mobile phone usage. In: CHI’11 Extended Abstracts on Human Factors in Com puting Systems. pp. 2191–2196. ACM (2011)

    work page 2011

  19. [19]

    In: Proceedings of the 2017 CHI Co nference on Hu- man Factors in Computing Systems

    Orji, R., Nacke, L.E., Di Marco, C.: Towards personality -driven persuasive health games and gamified systems. In: Proceedings of the 2017 CHI Co nference on Hu- man Factors in Computing Systems. pp. 1015–1027. ACM (2017)

    work page 2017

  20. [20]

    ACM Transactions on the Web (TWEB) 12(2), 9 (2018)

    Park, S., Matic, A., Garg, K., Oliver, N.: When simpler da ta does not imply less information: a study of user profiling scenarios with constr ained view of mobile http (s) traffic. ACM Transactions on the Web (TWEB) 12(2), 9 (2018)

    work page 2018

  21. [21]

    In: IFIP Conference on Human- Computer Interaction

    Raber, F., Krueger, A.: Towards understanding the influe nce of personality on mo- bile app permission settings. In: IFIP Conference on Human- Computer Interaction. pp. 62–82. Springer (2017)

    work page 2017

  22. [22]

    Journal of social and clinical psycholo gy 29(1), 95–122 (2010)

    Ryan, R.M., Bernstein, J.H., Brown, K.W.: Weekends, wor k, and well-being: Psy- chological need satisfactions and day of the week effects on m ood, vitality, and physical symptoms. Journal of social and clinical psycholo gy 29(1), 95–122 (2010)

    work page 2010

  23. [23]

    In: Proceed ings of the eighth symposium on usable privacy and security

    Staddon, J., Huffaker, D., Brown, L., Sedley, A.: Are priv acy concerns a turn- off?: engagement and privacy in social networks. In: Proceed ings of the eighth symposium on usable privacy and security. p. 10. ACM (2012)

    work page 2012

  24. [24]

    In: Proceedings of the 2012 ACM conference on ubiquitous computing

    Staiano, J., Lepri, B., Aharony, N., Pianesi, F., Sebe, N ., Pentland, A.: Friends don’t lie: inferring personality traits from social networ k structure. In: Proceedings of the 2012 ACM conference on ubiquitous computing. pp. 321– 330. ACM (2012)

    work page 2012

  25. [25]

    Tene, O., Polonetsky, J.: Big data for all: Privacy and us er control in the age of analytics. Nw. J. Tech. & Intell. Prop. 11, xxvii (2012)

    work page 2012

  26. [26]

    IEEE Trans- actions on Affective Computing 5(3), 273–291 (2014)

    Vinciarelli, A., Mohammadi, G.: A survey of personality computing. IEEE Trans- actions on Affective Computing 5(3), 273–291 (2014)

    work page 2014

  27. [27]

    Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Tec hnologies 2(3), 141 (2018)

    Wang, W., et al.: Sensing behavioral change over time: Us ing within-person vari- ability features from mobile sensing to predict personalit y traits. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Tec hnologies 2(3), 141 (2018)

    work page 2018

  28. [28]

    Computers in Human Behavior 62, 244–256 (2016)

    Xu, R., Frey, R.M., Fleisch, E., Ilic, A.: Understanding the impact of personality traits on mobile app adoption–insights from a large-scale fi eld study. Computers in Human Behavior 62, 244–256 (2016)

    work page 2016

  29. [29]

    In: Proceedings of the SIGCHI Con ference on Human Factors in Computing Systems

    Yee, N., et al.: Introverted elves & conscientious gnome s: the expression of person- ality in world of warcraft. In: Proceedings of the SIGCHI Con ference on Human Factors in Computing Systems. pp. 753–762. ACM (2011)

    work page 2011

  30. [30]

    In: Time use research in the social sciences, pp

    Zuzanek, J., Smale, B.J.: Life-cycle and across-the-we ek allocation of time to daily activities. In: Time use research in the social sciences, pp . 127–153. Springer (2002)

    work page 2002