Modeling Stress with Social Media Around Incidents of Gun Violence on College Campuses

Abstract

Stress constitutes a persistent wellbeing challenge to college students, impacting their personal, social, and academic life. However, violent events on campuses may aggravate student stress, due to the induced fear and trauma. In this paper, leveraging social media as a passive sensor of stress, we propose novel computational techniques to quantify and examine stress responses after gun violence on college campuses. We first present a machine learning classifier for inferring stress expression in Reddit posts, which achieves an accuracy of 82%. Next, focusing on 12 incidents of campus gun violence in the past five years, and social media data gathered from college Reddit communities, our methods reveal amplified stress levels following the violent incidents, which deviate from usual stress patterns on the campuses. Further, distinctive temporal and linguistic changes characterize the campus populations, such as reduced cognition, higher self pre-occupation and death-related conversations. We discuss the implications of our work in improving mental wellbeing and rehabilitation efforts around crisis events in college student populations.

Student Perspectives on Digital Phenotyping

“…In this paper we are interested in an emerging form of health surveillance technology, sometimes referred to as Digital Phenotyping, that uses passive sensing to identify and monitor problems. Data is collected via smartphones [60,84], social media [5,70], wearables [52], eLearning platforms [67] and more, which may then be useful for: (i) Monitoring students known to be at risk or with pre-diagnosed disorders and self reported problems; (ii) Monitoring all students to identify individuals who may be at risk and requiring help; (iii) Monitoring the student body as a whole in order to measure mental health and wellbeing and inform policy and Table 2: Self reports previously used in digital phenotyping research…”

Machine Learning in Mental Health

“…Here, textmining approaches were used [ 89 ] or proposed [ 28 ] to identify helpful and unhelpful comments in online mental health communities to assist human moderators to prioritize their responses to comments [ 28 , 89 ]. Saha and De Choudhury [ 165 ] further developed a classifier for inferring expressions of stress from Reddit posts by college students before and after incidences of gun violence; while others extracted linguistic features and topics in mental health communities to learn more about themes discussed online [ 133 , 141 ].…”

Machine learning in mental health: A systematic scoping review of methods and applications

“…Social media data also enables researchers to assess the impact of an incident on population mental health (e.g., stress levels of college students after experiencing gun violence [270]), and to track public response to disaster situations to inform the allocation of support resources [266,267,269]. ML applied to electronic health records was demonstrated to predict suicide risk with an accuracy similar to clinician assessment [268,271], as well as predict dementia and its risk factors with high accuracy [272].…”

“…Regression [281,282], DT [282], SVM [282], RF [281], Super Learner [281] Interview [282], Survey [281] Psychiatric Emergency BN [269], DT [269], SVM [269] Social Media [269] Psychiatric Stressors NLP [283], Named-entity recognition [283] Clinical Notes [283] Psychosis Regression [284], RF [285] Clinical Assessment [285], Electronic Health Records [284] Social Support LIWC [267], SVM [267] Social Media [267] Stress Cluster analysis [286], Sentiment Analysis [270], SVM [270] Clinical Assessment [286], Social Media [270] Substance Use PCA [265], NLP [265], RF [285] Social Media [265], Clinical Assessment [285] Suicide/Self-Harm ARM [271], DT [271], Genetic Algorithm [287], NB [268,271], RF [268,271], Regression [268,271,[288][289][290], SVM …”