STM Journals Peer-reviewed research from Consortium e-Learning Network

International Journal of Behavioral Sciences Review Article

Passive Digital Phenotyping for Longitudinal Burnout and Occupational Mental Health Surveillance: A Transformer-Based Explainable Deep Learning Approach Using Smartphone Behavioral Streams

  1. Hardik Srivastava Department of CSE, Greater Noida Institute of Technology, Greater Noida
  2. Shiraj Ahmad Department of CSE, Greater Noida Institute of Technology, Greater Noida

Abstract

Occupational burnout constitutes a pervasive yet chronically under-surveilled public health threat, its insidious temporal evolution rendering episodic self-report instruments structurally inadequate for early detection. This paper introduces BurnoutSense, a passive digital phenotyping framework that continuously harvests eight heterogeneous smartphone behavioral data streams encompassing application usage ecology, communication metadata, geospatial mobility, screen interaction dynamics, inferred sleep rhythmicity, keystroke kinematics, ambient noise exposure, and battery/charging cadence to construct individualized multivariate behavioral signatures sensitive to occupational mental health deterioration. At the computational core, we design the Hierarchical Temporal Transformer (HTT), a novel architecture incorporating irregular-sampling-aware positional encodings, circadian attention masking, and a dual-resolution temporal attention mechanism that simultaneously captures intra-day behavioral variability and inter-week longitudinal drift. To bridge the clinical explainability gap, we introduce the SHAP-Temporal Attribution Module (SHAP-TAM), which decomposes burnout risk predictions along both behavioral-feature and temporal-position axes, generating clinician-interpretable surveillance narratives. A 24-week longitudinal cohort study involving 387 knowledge-sector workers demonstrates that HTT achieves an Area Under the ROC Curve (AUC) of 0.931 for burnout onset prediction, with sensitivity 0.887 and specificity 0.906, significantly surpassing LSTM (AUC 0.847), Bi-GRU (AUC 0.863), Informer (AUC 0.879), and standard Transformer (AUC 0.891) baselines. The framework integrates differential privacy (epsilon = 0.8, delta = 10^-5) and federated inference to ensure sensitive behavioral data never leaves the user's device in identifiable form. BurnoutSense pioneers a clinically actionable, ethically grounded, always-on paradigm for occupational mental health surveillance.

Keywords

References (31)

  1. World Health Organization, "Burn-out an 'occupational phenomenon': International Classification of Diseases," WHO Technical Report, Geneva, Switzerland, 2019.
  2. Koeske GF, Koeske RD. Construct validity of the Maslach Burnout Inventory: A critical review and reconceptualization. The Journal of Applied Behavioral Science. 1989 May;25(2):131-44.
  3. Kristensen TS, Borritz M, Villadsen E, Christensen KB. The Copenhagen Burnout Inventory: A new tool for the assessment of burnout. Work & stress. 2005 Jul 1;19(3):192-207.
  4. Lathia N, Sandstrom GM, Mascolo C, Rentfrow PJ. Happier people live more active lives: Using smartphones to link happiness and physical activity. PloS one. 2017 Jan 4;12(1): e0160589.
  5. Bladon S, Eisner E, Bucci S, Oluwatayo A, Martin GP, Sperrin M, Ainsworth J, Faulkner S. A systematic review of passive data for remote monitoring in psychosis and schizophrenia. NPJ Digital Medicine. 2025 Jan 27;8(1):62.
  6. Sameh A, Rostami M, Oussalah M, Korpelainen R, Farrahi V. Digital phenotypes and digital biomarkers for health and diseases: a systematic review of machine learning approaches utilizing passive non-invasive signals collected via wearable devices and smartphones. Artificial Intelligence Review. 2024 Dec 21;58(2):66.
  7. Barnett I, Torous J, Staples P, Sandoval L, Keshavan M, Onnela JP. Relapse prediction in schizophrenia through digital phenotyping: a pilot study. Neuropsychopharmacology. 2018 Jul;43(8):1660-6.
  8. Torous J, Kiang MV, Lorme J, Onnela JP. New tools for new research in psychiatry: a scalable and customizable platform to empower data driven smartphone research. JMIR mental health. 2016 May 5;3(2): e5165.
  9. Wang R, Chen F, Chen Z, Li T, Harari G, Tignor S, Zhou X, Ben-Zeev D, Campbell AT. StudentLife: assessing mental health, academic performance and behavioral trends of college students using smartphones. In Proceedings of the 2014 ACM international joint conference on pervasive and ubiquitous computing 2014 Sep 13 (pp. 3-14).
  10. Saeb S, Zhang M, Karr CJ, Schueller SM, Corden ME, Kording KP, Mohr DC. Mobile phone sensor correlates of depressive symptom severity in daily-life behavior: an exploratory study. Journal of medical Internet research. 2015 Jul 15;17(7): e4273.
  11. Mehrotra A, Hendley R, Musolesi M. Towards multi-modal anticipatory monitoring of depressive states through the analysis of human-smartphone interaction. In Proceedings of the 2016 ACM international joint conference on pervasive and ubiquitous computing: adjunct 2016 Sep 12 (pp. 1132-1138).
  12. LiKamWa R, Liu Y, Lane ND, Zhong L. Moodscope: Building a mood sensor from smartphone usage patterns. InProceeding of the 11th annual international conference on Mobile systems, applications, and services 2013 Jun 25 (pp. 389-402).
  13. Rachuri KK, Musolesi M, Mascolo C, Rentfrow PJ, Longworth C, Aucinas A. EmotionSense: a mobile phones based adaptive platform for experimental social psychology research. In Proceedings of the 12th ACM international conference on Ubiquitous computing 2010 Sep 26 (pp. 281-290).
  14. Teh SK, Rawtaer I, Tan HP. Predictive accuracy of digital biomarker technologies for detection of mild cognitive impairment and pre-frailty amongst older adults: a systematic review and meta-analysis. IEEE Journal of Biomedical and Health Informatics. 2022 Jun 23;26(8):3638-48.
  15. Ha U, Madani S, Adib F. WiStress: Contactless stress monitoring using wireless signals. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies. 2021 Sep 14;5(3):1-37.
  16. Rana R, Higgins N, Stedman T, March S, Gucciardi DF, Barua PD, Joshi R. Passive AI detection of stress and burnout among frontline workers. Nursing Reports. 2025 Oct 22;15(11):373.
  17. Yuan Z, Cao Y, Chen Z. Smartphone-Based Psychological Sensing: A Large-Scale Study on the Impact of Extreme Isolation. Information Technology and Control. 2024 Dec 21;53(4):1152-68.
  18. Sano A, Picard RW. Stress recognition using wearable sensors and mobile phones. In2013 Humaine association conference on affective computing and intelligent interaction 2013 Sep 2 (pp. 671-676). IEEE.
  19. Dabre R, Fujita A. Recurrent stacking of layers for compact neural machine translation models. In Proceedings of the AAAI Conference on Artificial Intelligence 2019 Jul 17 (Vol. 33, No. 01, pp. 6292-6299).
  20. Fraikin A, Bennetot A, Allassonnière S. T-rep: Representation learning for time series using time-embeddings. In International Conference on Learning Representations 2024 May 31 (Vol. 2024, pp. 31285-31305).
  21. Song X, Deng L, Wang H, Zhang Y, He Y, Cao W. Deep learning-based time series forecasting. Artificial Intelligence Review. 2024 Nov 25;58(1):23.
  22. Chen W, Wang H, Zhang L, Zhang M. Temporal and spatial self supervised learning methods for electrocardiograms. Scientific Reports. 2025 Feb 19;15(1):6029.
  23. Chen K, Zhang D, Yao L, Guo B, Yu Z, Liu Y. Deep learning for sensor-based human activity recognition: Overview, challenges, and opportunities. ACM Computing Surveys (CSUR). 2021 May 22;54(4):1-40.
  24. Ribeiro MT, Singh S, Guestrin C. " Why should I trust you?" Explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining 2016 Aug 13 (pp. 1135-1144).
  25. Merz M, Richman R, Tsanakas A, Wüthrich MV. Interpreting deep learning models with marginal attribution by conditioning on quantiles. Data Mining and Knowledge Discovery. 2022 Jul;36(4):1335-70.
  26. Fournier Q, Caron GM, Aloise D. A practical survey on faster and lighter transformers. ACM Computing Surveys. 2023 Jul 17;55(14s):1-40.
  27. Golan T, Siegelman M, Kriegeskorte N, Baldassano C. Testing the limits of natural language models for predicting human language judgements. Nature Machine Intelligence. 2023 Sep;5(9):952-64.
  28. Zini JE, Awad M. On the explainability of natural language processing deep models. ACM Computing Surveys. 2022 Dec 3;55(5):1-31.
  29. Maslach C, Leiter MP. Early predictors of job burnout and engagement. Journal of applied psychology. 2008 May;93(3):498.
  30. Sadeghniiat-Haghighi K, Yazdi Z. Fatigue management in the workplace. Industrial psychiatry journal. 2015 Jan 1;24(1):12-7.
  31. Niedhammer I, Chastang JF, David S, Barouhiel L, Barrandon G. Psychosocial work environment and mental health: Job-strain and effort-reward imbalance models in a context of major organizational changes. International journal of occupational and environmental health. 2006 Apr 1;12(2):111-9.