StressMon

Abstract

Stress and depression are a common affliction in all walks of life. When left unmanaged, stress can inhibit productivity or cause depression. Depression can occur independently of stress. There has been a sharp rise in mobile health initiatives to monitor stress and depression. However, these initiatives usually require users to install dedicated apps or multiple sensors, making such solutions hard to scale. Moreover, they emphasise sensing individual factors and overlook social interactions, which plays a significant role in influencing stress and depression while being a part of a social system. We present StressMon, a stress and depression detection system that leverages single-attribute location data, passively sensed from the WiFi infrastructure. Using the location data, it extracts a detailed set of movement, and physical group interaction pattern features without requiring explicit user actions or software installation on client devices. These features are used in two different machine learning models to detect stress and depression. To validate StressMon, we conducted three different longitudinal studies at a university with different groups of students, totalling up to 108 participants. Our evaluation demonstrated StressMon detecting severely stressed students with a 96.01% True Positive Rate (TPR), an 80.76% True Negative Rate (TNR), and a 0.97 area under the ROC curve (AUC) score (a score of 1 indicates a perfect binary classifier) using a 6-day prediction window. In addition, StressMon was able to detect depression at 91.21% TPR, 66.71% TNR, and 0.88 AUC using a 15-day window. We end by discussing how StressMon can expand CSCW research, especially in areas involving collaborative practices for mental health management.

Machine Learning in Mental Health

“…Many of these works focus on the (early) detection (and monitoring) of depression or its symptoms (n = 10) [ 31 , 33 , 44 , 57 , 62 , 122 , 136 154 , 211 , 222 ], most often through the analysis of acoustic features of speech [ 31 , 122 ] or Twitter tweets [ 33 , 86 , 211 ]. Other examples include the detection of mood states from mobile sensing data [ 128 , 176 ], or phone typing dynamics [ 27 ] as well as stress assessments from location [ 218 ], biometrical and accelerometer data [ 67 ]. This is complemented by recent trends in analyzing human-robot [ 154 ] or agent interactions [ 155 ] to help assess peoples' mental health status.…”

“…, [ 4 , 50 , 136 ]), and discussed in terms of pragmatic challenges (e.g . , requirements of keeping technology charged and used; and users' compliant with data collection [ 44 , 152 , 153 , 176 , 201 ]; software compatibility issues in data extraction from varied devices [ 44 , 134 ]; and other technology infrastructure challenges [ 218 ]). Only a few studies described the active involvement of target-users, MHPs, or other domain experts in data labelling (e.g .…”

“…For example, Chang et al [ 31 ] adapted contextual inquiry [ 17 ] as a method to capture tacit diagnostic knowledge of MHPs in categorizing voice utterances of people suffering from major depressive disorder (MDD). Zakaria et al [ 218 ] conducted semi-structured interviews and collected survey data during their data collection study to "verify" primary causes of student stress. This includes information about how the students were managing their stress experiences, and insights about their work meeting dates, duration, and location.…”

“…Cao et al [ 27 ]; Mitra et al; Morshed et al [ 128 ]; Rabbi et al [ 153 ]; Salekin et al [ 168 ]; Zakaria et al [ 218 ] Confidential treatment/no (public) sharing of data Salekin et al [ 168 ]; Wang et al [ 201 ]; Zakaria et al [ 218 ]; Zhou et al [ 222 ] Consent and user control over data use; ability to contest ML Informed consent prior to study for primary data collection…”

“…Study risk assessment/planning with, or supervision by MHP (e.g ., licensed clinical psychologist, practicing psychiatrists) Nobles et al [ 134 ]; Salekin et al [ 168 ]; Zakaria et al [ 218 ] Study coordination by person trained with relevant expertise…”