. Author manuscript; available in PMC: 2020 Apr 21.

Published in final edited form as: Proc ACM Interact Mob Wearable Ubiquitous Technol. 2018 Dec;2(4):153. doi: 10.1145/3287031

Table 1.

Features extracted from the app interaction data.

Interaction Type	Features	Explanation
Single instance	Interaction length	The average length (in secs) of single instance interactions with the app.
Single instance	Interaction type (5 types based on the length of interaction)	The fraction of interactions labeled as Glances, Checks, Brief review, Detailed review and Deep review.
Weekly instance	Number of weekly interactions with the app	The average number of weekly interactions made by the participants in each behavioral cluster.
	Number of use days per week	The average number of days per week where the app was used by the participants in each behavioral cluster.
	Density of usage day groups divided into 3 categories: 0–1 days, 2–4 days, and 5–7 days	The fraction of weekly instances in a behavioral cluster where no. of use days is either 0–1, 2–4 or 5–7.