Exploring Unique App Signature of the Depressed and Non-depressed Through Their Fingerprints on Apps

Abstract

Growing research on re-identification through app usage behavior reveals the privacy threat in having smartphone usage data to third parties. However, re-identifiability of a vulnerable group like the depressed is unexplored. We fill this knowledge gap through an in the wild study on 100 students' PHQ-9 scale's data and 7 days' logged app usage data. We quantify the uniqueness and re-identifiability through exploration of minimum hamming distance in terms of the set of used apps. Our findings show that using app usage data, each of the depressed and non-depressed students is re-identifiable. In fact, using only 7 h' data of a week, on average, 91% of the depressed and 88% of the non-depressed are re-identifiable. Moreover, data of a single app category (i.e., Tools) can also be used to re-identify each depressed student. Furthermore, we find that the rate of uniqueness among the depressed students is significantly higher in some app categories. For instance, in the Social Media category, the rate of uniqueness is 9% higher (P = .02, Cohen's d = 1.31) and in the Health & Fitness category, this rate is 8% higher (P = .005, Cohen's d = 1.47) than the non-depressed group. Our findings suggest that each of the depressed students has a unique app signature which makes them re-identifiable. Therefore, during the design of the privacy protecting systems, designers need to consider the uniqueness of them to ensure better privacy for this vulnerable group.

Less Is More: Leveraging Digital Behavioral Markers for Real-Time Identification of Loneliness in Resource-Limited Settings

“…SHAP analysis [4] on our best ML model NB based on the IG selected 17 features showed that the lonely students were more likely to have a higher variation in duration per app on weekdays over the 4 time periods. This finding aligns with the studies conducted through conventional statistical methods showing depressed students' variation of diurnal app usage patterns [17,38]. The variation can reflect students' mood swings while going through a negative experience [24].…”

“…However, students used several apps which were not available there, and thus, using the package name, we explored the features of those apps in online app stores (e.g., apkmonk.com) and developers' websites. Finally, we categorized the apps by understanding the app categorization process of previous studies (e.g., [17]) and through a discussion with 2 graduate students of the engineering faculty. We found 105 students using 867 apps of 26 categories (Fig.…”

“…Ratio of hamming distance. As uniqueness in terms of app usage varies among smartphone users (e.g., between the depressed and non-depressed [17]), in the case of each student, we calculated the ratio of the hamming distance from the nearest lonely to the nearest non-lonely student. To get an unbiased ML model, we did not consider the group (e.g., non-lonely) of that student while calculating the distance: 𝐷𝐿 𝑖𝑗 = (𝐴 𝑖 ∪ 𝐴 𝑗 ) − (𝐴 𝑖 ∩ 𝐴 𝑗 ) where 𝐷𝐿 𝑖𝑗 denotes the distance of the 𝑖 𝑡ℎ student from the 𝑗 𝑡ℎ student of the lonely group; 𝐴 𝑖 and 𝐴 𝑗 denote the set of apps used by the 𝑖 𝑡ℎ and 𝑗 𝑡ℎ students respectively.…”

“…Other behavioral data: For total smartphone usage (i.e., regardless of app category) and each of the 26 app categories, we also extracted duration, frequency of launch, duration per launch, launch per app, duration per app, and session data to extract the participants' app usage patterns. In addition, as app usage behavior varies by days [17], we extracted features regarding the difference in app usage between weekdays and weekends dividing the days (Weekdays: Sunday to Thursday, Weekends: Friday and Saturday) based on the working schedule of Bangladesh.…”

A Minimal and Faster System to Identify Depression Through Smartphone: An Explainable Machine Learning-Based Approach

“…Smartphones have become affordable [82] and are available to the majority of adults in emerging and developing countries [16]. Smartphone usage has a significant relation with depression [27,28,44,45] and loneliness [40,41,42]. In addition, higher levels of depression are linked with higher usage and a higher number of times interacting on the phone [44].…”