Predicting Human Depression with Hybrid Data Acquisition utilizing Physical Activity Sensing and Social Media Feeds

2025-05-28

AI Summary

This study addresses the challenge of non-invasive depression severity assessment by proposing a multimodal predictive framework integrating smartphone-based physical activity sensing with Twitter text sentiment analysis. Methodologically, it introduces a novel joint modeling approach combining six-dimensional movement features (e.g., step-frequency variability, activity rhythmicity) and three-dimensional social-affective features (e.g., loneliness-related word frequency, interaction sparsity). Convolutional neural networks (CNNs) model activity patterns, naïve Bayes classifiers analyze sentiment polarity, and support vector machines (SVMs) perform depression severity classification. Evaluated on a cohort of 33 participants, the framework achieves 95.0% activity recognition accuracy, 95.6% sentiment analysis accuracy, and 94.0% depression severity classification accuracy—significantly outperforming MLP and KNN baselines—and is clinically calibrated against the Geriatric Depression Scale (GDS). This work is the first to uncover cross-modal depressive associations between daily motor rhythmicity and social withdrawal behaviors, offering a deployable, privacy-preserving paradigm for digital mental health intervention.

Technology Category

• Machine Learning: Multimodal Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Planning, Routing, and Scheduling: Activity and Plan Recognition

Application Category

• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• Security and Privacy: Large-scale security measurements
• User Modeling, Personalization and Recommendation: Practical large-scale studies of user experience

Abstract

Mental disorders including depression, anxiety, and other neurological disorders pose a significant global challenge, particularly among individuals exhibiting social avoidance tendencies. This study proposes a hybrid approach by leveraging smartphone sensor data measuring daily physical activities and analyzing their social media (Twitter) interactions for evaluating an individual's depression level. Using CNN-based deep learning models and Naive Bayes classification, we identify human physical activities accurately and also classify the user sentiments. A total of 33 participants were recruited for data acquisition, and nine relevant features were extracted from the physical activities and analyzed with their weekly depression scores, evaluated using the Geriatric Depression Scale (GDS) questionnaire. Of the nine features, six are derived from physical activities, achieving an activity recognition accuracy of 95%, while three features stem from sentiment analysis of Twitter activities, yielding a sentiment analysis accuracy of 95.6%. Notably, several physical activity features exhibited significant correlations with the severity of depression symptoms. For classifying the depression severity, a support vector machine (SVM)-based algorithm is employed that demonstrated a very high accuracy of 94%, outperforming alternative models, e.g., the multilayer perceptron (MLP) and k-nearest neighbor. It is a simple approach yet highly effective in the long run for monitoring depression without breaching personal privacy.

Problem

• Predicting depression using physical activity and social media data
• Hybrid approach combining sensor data and sentiment analysis
• High-accuracy SVM model for depression severity classification

Innovation

• Hybrid data: physical activity and social media
• CNN and Naive Bayes for activity and sentiment
• SVM achieves 94% depression classification accuracy