Unobtrusive stress detection on the basis of smartphone usage data

Elena Vildjiounaite (Corresponding Author), Johanna Kallio, Vesa Kyllönen, Mikko Nieminen, Ilmari Määttänen, Mikko Lindholm, Jani Mäntyjärvi, Georgy Gimel'farb

Research output: Contribution to journalArticleScientificpeer-review

4 Citations (Scopus)

Abstract

Stress has become an important health problem, but existing stress detectors are inconvenient in long-term real-life use because users either have to wear dedicated devices or expend notable interaction efforts in system adaptation to specifics of each person. Adaptation is necessary because individuals significantly differ in their perception of stress and stress responses, but typical adaptation employs supervised learning methods and hence requires fairly large sets of labelled data (i.e. information on whether each reporting period was stressful or not) from every user. To address these problems, we propose a novel unsupervised stress detector, based on using a smartphone as the only device and using discrete hidden Markov models (HMM) with maximum posterior marginal (MPM) decisions for analysis of phone data. Our detector requires neither additional hardware nor data labelling and hence is truly unobtrusive and suitable for lifelong use. Its accuracy was evaluated using two real-life datasets: in the first case, adaptation was based on very short (a few days) phone interaction histories of each individual, and in the second case—on longer histories. In these tests, the proposed HMM-MPM achieved 59 and 70% accuracies, respectively, which is comparable with results of fully supervised methods, reported by other works.
Original languageEnglish
Pages (from-to)671-688
Number of pages18
JournalPersonal and Ubiquitous Computing
Volume22
Issue number4
DOIs
Publication statusPublished - 17 Jan 2018
MoE publication typeA1 Journal article-refereed

Fingerprint

Smartphones
Hidden Markov models
Detectors
Supervised learning
Medical problems
Labeling
Wear of materials
Hardware
Interaction
Hidden Markov model

Keywords

  • Mobile phone data analysis
  • Stress detection
  • Personalisation
  • Unsupervised learning
  • Hidden Markov models

Cite this

@article{58cb15878efc418abe0d17e397a1da23,
title = "Unobtrusive stress detection on the basis of smartphone usage data",
abstract = "Stress has become an important health problem, but existing stress detectors are inconvenient in long-term real-life use because users either have to wear dedicated devices or expend notable interaction efforts in system adaptation to specifics of each person. Adaptation is necessary because individuals significantly differ in their perception of stress and stress responses, but typical adaptation employs supervised learning methods and hence requires fairly large sets of labelled data (i.e. information on whether each reporting period was stressful or not) from every user. To address these problems, we propose a novel unsupervised stress detector, based on using a smartphone as the only device and using discrete hidden Markov models (HMM) with maximum posterior marginal (MPM) decisions for analysis of phone data. Our detector requires neither additional hardware nor data labelling and hence is truly unobtrusive and suitable for lifelong use. Its accuracy was evaluated using two real-life datasets: in the first case, adaptation was based on very short (a few days) phone interaction histories of each individual, and in the second case—on longer histories. In these tests, the proposed HMM-MPM achieved 59 and 70{\%} accuracies, respectively, which is comparable with results of fully supervised methods, reported by other works.",
keywords = "Mobile phone data analysis, Stress detection, Personalisation, Unsupervised learning, Hidden Markov models",
author = "Elena Vildjiounaite and Johanna Kallio and Vesa Kyll{\"o}nen and Mikko Nieminen and Ilmari M{\"a}{\"a}tt{\"a}nen and Mikko Lindholm and Jani M{\"a}ntyj{\"a}rvi and Georgy Gimel'farb",
note = "Two missing autors added as per published corregendum.",
year = "2018",
month = "1",
day = "17",
doi = "10.1007/s00779-017-1108-z",
language = "English",
volume = "22",
pages = "671--688",
journal = "Personal and Ubiquitous Computing",
issn = "1617-4909",
publisher = "Springer",
number = "4",

}

Unobtrusive stress detection on the basis of smartphone usage data. / Vildjiounaite, Elena (Corresponding Author); Kallio, Johanna; Kyllönen, Vesa; Nieminen, Mikko; Määttänen, Ilmari; Lindholm, Mikko; Mäntyjärvi, Jani; Gimel'farb, Georgy.

In: Personal and Ubiquitous Computing, Vol. 22, No. 4, 17.01.2018, p. 671-688.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Unobtrusive stress detection on the basis of smartphone usage data

AU - Vildjiounaite, Elena

AU - Kallio, Johanna

AU - Kyllönen, Vesa

AU - Nieminen, Mikko

AU - Määttänen, Ilmari

AU - Lindholm, Mikko

AU - Mäntyjärvi, Jani

AU - Gimel'farb, Georgy

N1 - Two missing autors added as per published corregendum.

PY - 2018/1/17

Y1 - 2018/1/17

N2 - Stress has become an important health problem, but existing stress detectors are inconvenient in long-term real-life use because users either have to wear dedicated devices or expend notable interaction efforts in system adaptation to specifics of each person. Adaptation is necessary because individuals significantly differ in their perception of stress and stress responses, but typical adaptation employs supervised learning methods and hence requires fairly large sets of labelled data (i.e. information on whether each reporting period was stressful or not) from every user. To address these problems, we propose a novel unsupervised stress detector, based on using a smartphone as the only device and using discrete hidden Markov models (HMM) with maximum posterior marginal (MPM) decisions for analysis of phone data. Our detector requires neither additional hardware nor data labelling and hence is truly unobtrusive and suitable for lifelong use. Its accuracy was evaluated using two real-life datasets: in the first case, adaptation was based on very short (a few days) phone interaction histories of each individual, and in the second case—on longer histories. In these tests, the proposed HMM-MPM achieved 59 and 70% accuracies, respectively, which is comparable with results of fully supervised methods, reported by other works.

AB - Stress has become an important health problem, but existing stress detectors are inconvenient in long-term real-life use because users either have to wear dedicated devices or expend notable interaction efforts in system adaptation to specifics of each person. Adaptation is necessary because individuals significantly differ in their perception of stress and stress responses, but typical adaptation employs supervised learning methods and hence requires fairly large sets of labelled data (i.e. information on whether each reporting period was stressful or not) from every user. To address these problems, we propose a novel unsupervised stress detector, based on using a smartphone as the only device and using discrete hidden Markov models (HMM) with maximum posterior marginal (MPM) decisions for analysis of phone data. Our detector requires neither additional hardware nor data labelling and hence is truly unobtrusive and suitable for lifelong use. Its accuracy was evaluated using two real-life datasets: in the first case, adaptation was based on very short (a few days) phone interaction histories of each individual, and in the second case—on longer histories. In these tests, the proposed HMM-MPM achieved 59 and 70% accuracies, respectively, which is comparable with results of fully supervised methods, reported by other works.

KW - Mobile phone data analysis

KW - Stress detection

KW - Personalisation

KW - Unsupervised learning

KW - Hidden Markov models

UR - http://www.scopus.com/inward/record.url?scp=85040704304&partnerID=8YFLogxK

U2 - 10.1007/s00779-017-1108-z

DO - 10.1007/s00779-017-1108-z

M3 - Article

VL - 22

SP - 671

EP - 688

JO - Personal and Ubiquitous Computing

JF - Personal and Ubiquitous Computing

SN - 1617-4909

IS - 4

ER -