CoBAD: Modeling Collective Behaviors for Human Mobility Anomaly Detection

2025-08-13

AI Summary

This paper addresses the underexplored challenge of collective anomaly detection in human mobility. Unlike conventional individual-level anomaly detection, it formally defines and identifies two novel types of collective anomalies—unexpected co-occurrence and absence anomalies—thereby filling a critical gap in absence-anomaly research. Methodologically, the approach constructs collective event sequences and co-occurrence event graphs, and introduces a two-stage attention mechanism to jointly model individual mobility patterns and inter-individual spatiotemporal interactions. It further employs a self-supervised pretraining strategy based on masked event prediction and link reconstruction, integrated with graph neural networks for end-to-end learning. Evaluated on large-scale real-world datasets, the proposed model achieves substantial improvements: +13–18% in AUC-ROC and +19–70% in AUC-PR over state-of-the-art baselines, demonstrating superior collective anomaly detection performance.

Technology Category

• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Humans and AI: Crowd Sourcing and Human Computation
• Machine Learning: Graph-based Machine Learning

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Security and Privacy: Large-scale security measurements

Abstract

Detecting anomalies in human mobility is essential for applications such as public safety and urban planning. While traditional anomaly detection methods primarily focus on individual movement patterns (e.g., a child should stay at home at night), collective anomaly detection aims to identify irregularities in collective mobility behaviors across individuals (e.g., a child is at home alone while the parents are elsewhere) and remains an underexplored challenge. Unlike individual anomalies, collective anomalies require modeling spatiotemporal dependencies between individuals, introducing additional complexity. To address this gap, we propose CoBAD, a novel model designed to capture Collective Behaviors for human mobility Anomaly Detection. We first formulate the problem as unsupervised learning over Collective Event Sequences (CES) with a co-occurrence event graph, where CES represents the event sequences of related individuals. CoBAD then employs a two-stage attention mechanism to model both the individual mobility patterns and the interactions across multiple individuals. Pre-trained on large-scale collective behavior data through masked event and link reconstruction tasks, CoBAD is able to detect two types of collective anomalies: unexpected co-occurrence anomalies and absence anomalies, the latter of which has been largely overlooked in prior work. Extensive experiments on large-scale mobility datasets demonstrate that CoBAD significantly outperforms existing anomaly detection baselines, achieving an improvement of 13%-18% in AUCROC and 19%-70% in AUCPR. All source code is available at https://github.com/wenhaomin/CoBAD.

Problem

• Detecting collective human mobility anomalies from spatiotemporal dependencies
• Modeling unexpected co-occurrence and absence anomalies in group behaviors
• Addressing underexplored challenges in unsupervised collective anomaly detection

Innovation

• Two-stage attention mechanism
• Unsupervised learning with co-occurrence graph
• Masked event and link reconstruction