Applications for Future Internet. International Summit, AFI 2016, Puebla, Mexico, May 25-28, 2016, Revised Selected Papers

Research Article

Using Intermediate Models and Knowledge Learning to Improve Stress Prediction

Download42 downloads
  • @INPROCEEDINGS{10.1007/978-3-319-49622-1_16,
        author={Alban Maxhuni and Pablo Hernandez-Leal and Eduardo Morales and L. Sucar and Venet Osmani and Angelica Muńoz-Mel\^{e}ndez and Oscar Mayora},
        title={Using Intermediate Models and Knowledge Learning to Improve Stress Prediction},
        proceedings={Applications for Future Internet. International Summit, AFI 2016, Puebla, Mexico, May 25-28, 2016, Revised Selected Papers},
        proceedings_a={AFI360},
        year={2017},
        month={1},
        keywords={Motor activity Stress prediction Smartphones},
        doi={10.1007/978-3-319-49622-1_16}
    }
    
  • Alban Maxhuni
    Pablo Hernandez-Leal
    Eduardo Morales
    L. Sucar
    Venet Osmani
    Angelica Muńoz-Meléndez
    Oscar Mayora
    Year: 2017
    Using Intermediate Models and Knowledge Learning to Improve Stress Prediction
    AFI360
    Springer
    DOI: 10.1007/978-3-319-49622-1_16
Alban Maxhuni1,*, Pablo Hernandez-Leal2,*, Eduardo Morales2,*, L. Sucar2,*, Venet Osmani3,*, Angelica Muńoz-Meléndez2,*, Oscar Mayora3,*
  • 1: University of Trento
  • 2: INAOE-Instituto Nacional de Astrofísica, Óptica y Electrónica
  • 3: CREATE-NET
*Contact email: maxhuni@disi.unitn.it, pablohl@inaoep.mx, emorales@inaoep.mx, esucar@inaoep.mx, vosmani@create-net.org, munoz@inaoep.mx, omayora@create-net.org

Abstract

Motor activity in physical and psychological stress exposure has been studied almost exclusively with self-assessment questionnaires and from reports that derive from human observer, such as verbal rating and simple descriptive scales. However, these methods are limited in objectively quantifying typical behaviour of stress. We propose to use accelerometer data from smartphones to objectively quantify stress levels. Used data was collected in real-world setting, from 29 employees in two different organisations over 5 weeks. To improve classification performance we propose to use . These intermediate models represent the mood state of a person which is used to build the final stress prediction model. In particular, we obtained an accuracy of 78.2 % to classify stress levels.