Explainable Detection of Depression Status Shifts from User Digital Traces

2026-05-14

AI Summary

This work addresses the dynamic identification of depression state trajectories—such as deterioration, improvement, or stability—from users’ digital traces. To this end, we propose an interpretable framework that first employs a multi-task BERT model to jointly extract signals related to sentiment, emotion, and depression severity, thereby constructing user-level temporal trajectories. Subsequently, change-point detection is integrated to capture critical transitions in these trajectories. Finally, a large language model generates human-readable evolution reports that exhibit high coverage, strong temporal coherence, and sensitivity to state shifts. Experiments on two social media datasets demonstrate that, compared to directly prompting a large language model, our approach produces summaries significantly superior in comprehensiveness, temporal logicality, and responsiveness to pivotal transition points.

Abstract

Every day, users generate digital traces (e.g., social media posts, chats, and online interactions) that are inherently timestamped and may reflect aspects of their mental state. These traces can be organized into temporal trajectories that capture how a user's mental health signals evolve, including phases of improvement, deterioration, or stability. In this work, we propose an explainable framework for detecting and analyzing depression-related status shifts in user digital traces. The approach combines multiple BERT-based models to extract complementary signals across different dimensions (e.g., sentiment, emotion, and depression severity). Such signals are then aggregated over time to construct user-level trajectories that are analyzed to identify meaningful change points. To enhance interpretability, the framework integrates a large language model to generate concise and human-readable reports that describe the evolution of mental-health signals and highlight key transitions. We evaluate the framework on two social media datasets. Results show that the approach produces more coherent and informative summaries than direct LLM-based reporting, achieving higher coverage of user history, stronger temporal coherence, and improved sensitivity to change points. An ablation study confirms the contribution of each component, particularly temporal modeling and segmentation. Overall, the method provides an interpretable view of mental health signals over time, supporting research and decision making without aiming at clinical diagnosis.

Problem

• depression status shifts
• digital traces
• explainable detection
• mental health trajectories
• temporal change points

Innovation

• explainable AI
• depression detection
• temporal trajectory
• BERT-based modeling
• large language model