Social Media Behaviour and Perceptions of AI-Assisted Mental Health Monitoring: A Survey-Based Machine Learning Study
Multimedia Research · 2026
DOI: 10.46253/j.mr.v9i4.a1 ↗Kurz gefasst
Fachartikel in der Zeitschrift Multimedia Research, 2026. Verfasst von Shweta Marigoudar und 1 weiteren Autor:innen.
Abstract (Originalsprache)
Passive, continuous monitoring of everyday digital behaviour is increasingly proposed as a complement to episodic clinical assessment of mental health, yet whether people accept such monitoring has received far less empirical attention than whether it is technically feasible.This study investigates both sides of that question in a single, survey-based design.A structured instrument covering demographics, social media usage, emotional expression, and attitudes toward AI-assisted mental health detection was administered through Google Forms to 132 respondents.After validation, cleaning, and label encoding of the categorical responses, an exploratory analysis surfaced behavioural and attitudinal patterns, and five supervised classifiers Logistic Regression, Decision Tree, Random Forest, Naïve Bayes, and Support Vector Machine (SVM) were trained and compared on the encoded, non-textual survey data using a 75/25 train-test split.Random Forest achieved the highest accuracy (50.0%), followed by SVM and Naïve Bayes (43.8% each), while Logistic Regression (25.0%) and Decision Tree (21.9%) performed the weakest; this ceiling is attributable to the small sample size, the purely categorical feature set, and pronounced class imbalance in the target attribute.Independently of classification performance, the survey revealed considerable openness toward AI-assisted mental health detection: 58% of respondents expressed confidence in such systems and 61% indicated willingness to use one, subject to privacy and transparency safeguards.These results provide a user-acceptance perspective that complements the predominantly text-mining orientation of existing mental health detection research.
Zitieren
APA
Marigoudar, S., & S, R. (2026). Social Media Behaviour and Perceptions of AI-Assisted Mental Health Monitoring: A Survey-Based Machine Learning Study. Multimedia Research. https://doi.org/10.46253/j.mr.v9i4.a1
Vancouver
Marigoudar S, S R. Social Media Behaviour and Perceptions of AI-Assisted Mental Health Monitoring: A Survey-Based Machine Learning Study. Multimedia Research. 2026. doi:10.46253/j.mr.v9i4.a1
BibTeX
@article{Marigoudar2026,
author = {Marigoudar, Shweta and S, Reethu},
title = {Social Media Behaviour and Perceptions of AI-Assisted Mental Health Monitoring: A Survey-Based Machine Learning Study},
journal = {Multimedia Research},
year = {2026},
doi = {10.46253/j.mr.v9i4.a1}
}
EUVIMED