Contact network

Key characteristics

• Topology: networks may be sparse or dense, clustered or disassortative; common measures include degree distribution, clustering coefficient, and path length.
• Edge attributes: contacts can be weighted (frequency, duration, or strength), directed (one-way exposure), or labelled by context (household, workplace, transit).
• Temporal structure: many contact networks are dynamic—contacts appear and disappear over time—so timing and concurrency strongly influence outcomes.
• Layers and multiplexity: individuals participate in multiple contexts (family, work, social), which can be modelled as overlapping layers.

Historically, the concept grew from classical graph theory and early sociometric studies; it became central to epidemiology with compartmental models that were extended to structured populations. Advances in sensors, mobile devices and data collection have expanded the ability to measure contact networks directly, enabling more realistic simulations and targeted interventions.

Uses and examples

• Modeling infectious disease transmission to evaluate vaccination, testing, or social-distancing strategies.
• Contact tracing: identifying recent contacts of a case to interrupt chains of transmission.
• Studying information diffusion or marketing cascades in social systems.
• Analyzing failure propagation in engineered systems or supply chains.

Construction of contact networks relies on surveys, diaries, proximity sensors (Bluetooth, RFID), and administrative records. Each method has trade-offs in coverage, temporal resolution and privacy risk. Analysts must account for sampling bias, missing links, and measurement error when drawing conclusions.

Distinguishing a contact network from related concepts helps clarity: it is a model of potential transmission pathways, not the same as a transmission tree (which records who actually infected whom), and it differs from abstract social networks that emphasize long-term ties rather than short-term proximate contacts. Contact networks remain a foundational tool for understanding and managing processes that depend on who meets whom.