Hyperbolic mapping of human proximity networks.

Area of Science:

• Complex systems science
• Network science
• Epidemiology

Background:

• Human proximity networks are crucial for understanding disease and information transmission.
• These temporal networks are often sparse in individual snapshots.
• Previous methods focused on traditional, non-temporal complex networks.

Purpose of the Study:

• To map human proximity networks into hyperbolic spaces.
• To investigate the utility of hyperbolic embeddings for analyzing temporal network properties.
• To explore applications in community detection, routing, and epidemic prediction.

Main Methods:

• Time-aggregated representation of human proximity networks.
• Embedding aggregated networks into hyperbolic space using established complex network methods.
• Theoretical justification and experimental validation across six real-world systems.

Main Results:

• Successful hyperbolic embedding of time-aggregated human proximity networks.
• Hyperbolic maps facilitate community identification and efficient greedy routing.
• Hyperbolic distance predicts epidemic arrival times and future links with high precision.

Conclusions:

• Hyperbolic embedding is a compatible and powerful method for analyzing human proximity networks.
• This approach enhances understanding of network structure, dynamics, and epidemic spreading.
• Hyperbolic mapping offers a novel perspective for predicting emergent behaviors in human systems.