Understanding and Predicting the Burst of Burnout via Social Media

Jue Wu, Junyi Ma, Yasha Wang, Jiangtao Wang

    Research output: Contribution to journalConference articlepeer-review

    9 Citations (Scopus)

    Abstract

    Job burnout is a special type of work-related stress that is prevalent in our modern society, and constant burnout is extremely harmful for people's physical health and emotional wellbeing. Traditional studies for burnout mainly rely on surveys/questionnaires, which have revealed several interesting findings but are of high cost and very time consuming. With the prevalence of social networking applications, we aim to re-investigate the burnout phenomenon in a novel perspective. In this paper, we collected a dataset consisting of 1532 burnout Weibo users with their postings. Based on the previous literature, we propose a number of hypotheses about what might be the burst signal of the burnout from the perspective of language, time and interaction. Furthermore, extensive correlation analysis is conducted to investigate if these hypotheses are supported, which leads to a number of interesting findings. Finally, we develop machine learning models to predict the burst of burnout based on extracted features and achieve a relatively high accuracy, which reveals potential implications in early-stage intervention.
    Original languageEnglish
    Article number265
    Pages (from-to)1-27
    Number of pages27
    JournalProceedings of the ACM on Human-Computer Interaction
    Volume4
    Issue numberCSCW3
    DOIs
    Publication statusPublished - 5 Jan 2021

    Keywords

    • Computer Networks and Communications
    • Human-Computer Interaction
    • Social Sciences (miscellaneous)

    Fingerprint

    Dive into the research topics of 'Understanding and Predicting the Burst of Burnout via Social Media'. Together they form a unique fingerprint.

    Cite this