Deriving College Students’ Phone Call Patterns to Improve Student Life

Abstract

When high school students leave their homes for a college education, the students often face enormous changes and challenges in life, such as meeting new people, more responsibilities in life, and being away from family and their comfort zones. These sudden changes often lead to an elevation of stress and anxiety, affecting a student's mental health and well-being and academic progress. With the outbreak of a pandemic, such as COVID-19, this transition moment can worsen with frequent, long-lasting lockdown and associated academic disruption. To help this young population, researchers are increasingly relying on smartphones and wearables, such as smartwatches, to continuously monitor students' daily lives to identify various factors that can affect students' phone call patterns associated with their health and well-being and academic success. In this work, we use different visualizations and statistical techniques to find various geographical places and temporal factors that affect students' phone call patterns (in terms of phone call duration and frequency) to foster the design and delivery of future smartphone-based health interventions using predictive models; thereby, potentially helping students adjust to college life with or without the presence of an emergency, such as pandemic that adversely impacts academic calendar and student life. From our detailed analysis of an 18-month dataset collected from a cohort of 464 freshmen, we obtain insights on communication pattern variations during different temporal contexts, e.g., epochs of a day, days of a week, the parts of a semester, social events, and in various geographical contexts (i.e., places of interest). Finally, we also obtain a negative correlation of −0.29 between physical activity and phone call duration, which can help provide guided feedback to improve future health behaviors.

PMSIEMDL: Design of a Pattern analysis Model for identification of Student Inclination towards different Educational-fields via Multimodal Deep Learning Fusions

“…Some models may help students behave better in certain situations. Student health-related concepts and improved real-time online learning performance are the focus of [33], an example of how mobile health, temporal parameters, and geographic features may be used to benefit students.…”

Environment Knowledge-Driven Generic Models to Detect Coughs From Audio Recordings

“…Also, sensors placed around the throat are not unobtrusive, and people may find it uncomfortable. Most importantly, wearable device user adherence drops over time [19] , [20] , [21] , [22] , [23] , [24] , [25] , [26] , [27] , [28] , [29] , [30] , [31] , [32] , [33] , [34] , [35] , [36] , [37] , [38] , [39] .…”