Weighted Network Propagation Model¶
Simulation or model — instantiates Heterogeneous Medium Propagation Routing
A graph model where nodes and ties have different propagation weights or receptivity values.
When the medium is a discrete network — people, computers, cities linked by edges — and the question is how far and how fast an effect spreads, you run it forward on the graph. A Weighted Network Propagation Model represents the medium as nodes and edges, gives each edge a transmission weight and each node a receptivity, and then simulates the propagating entity hopping across the graph step by step. Its defining move is dynamics on a discrete weighted graph: unlike a continuous cost surface, it works on explicit nodes and links, and unlike a static map, it steps a cascade forward — seeding some nodes, letting the effect jump along weighted edges, and reporting who is reached, how many, and when.
Example¶
A hospital epidemiology team wants to know how far an outbreak of a resistant bacterium will spread through the ward network and whether isolating one unit would blunt it. They build a weighted network model: nodes are patients, staff, and wards; edges carry transmission weights set by contact intensity — a nurse who moves between wards is a heavily weighted link, a patient in a single room a lightly connected node; each node has a susceptibility. Seeding the index case, they run the contagion forward across the graph.
The simulation shows the outbreak's shape over time: it smolders within the source ward, then leaps — via that high-weight roaming nurse — into two distant units, and the total reached hinges almost entirely on a few such high-transmission connectors. Re-running with the nurse's cross-ward edges cut (a simulated cohorting policy) collapses the projected spread. The model's payoff is not a static picture but a dynamic answer — reach, timing, and the leverage of specific connectors — that tells the team which single intervention matters most.
How it works¶
The model's identity is weighted graph in, cascade out:
- Build the weighted graph. Nodes get receptivity values, edges get transmission weights reflecting how strongly the medium conducts along each link — the discrete topology of the network, interfaces and all.
- Seed and step the dynamics. Starting from seed nodes, the effect propagates along edges each round with a probability or intensity set by the edge weight and the receiving node's susceptibility; the cascade grows, saturates, or dies out over time.
- Account for gain and loss along the way. Each hop attenuates or amplifies the effect according to edge weight and node state, so the model tracks where transmission builds into a runaway cascade and where it damps out — and which high-weight connectors carry most of the spread.
Where a continuous solver meshes a body and a routing simulation extracts a corridor, this model runs an explicit stochastic or deterministic cascade over a discrete weighted network.
Tuning parameters¶
- Edge-weight and receptivity calibration — how transmission strengths and node susceptibilities are set. These weights drive everything, and they are usually the least certain input.
- Propagation rule — the contagion dynamics chosen (simple threshold, independent-cascade, compartmental SIR-style). Each yields qualitatively different spread and fits different phenomena.
- Seeding — which and how many nodes start infected/activated, which strongly shapes reach and reveals the leverage of well-placed seeds.
- Stochastic vs. deterministic — whether the cascade is run as many random trials or a single averaged flow. Trials expose variance and tail risk; the average is cleaner but hides rare explosive outcomes.
- Time horizon — how many steps the cascade runs, setting whether you see the early phase or the saturated end state.
When it helps, and when it misleads¶
Its strength is answering the reach-and-timing question on a discrete medium — total spread, speed, and, most usefully, which nodes and edges are the leverage points where an intervention most changes the outcome. It naturally captures superspreading: because transmission is heterogeneous, a small number of high-weight connectors drive a disproportionate share of the cascade, so targeting them beats treating all nodes alike.[n1]
Its failure mode is over-trust in a graph and weights that are only approximately known: an omitted edge or a mis-set weight can flip a projection from contained to runaway, and the crisp output invites false precision over what is really a structured guess. Stochastic spread also has fat tails, so a single averaged run can badly understate the chance of a rare explosive cascade. The classic misuse is reading the model's numbers as forecasts rather than as comparisons between interventions. The guarding discipline is to run many stochastic trials and report the distribution, stress-test the weights, and use the model to rank interventions rather than to predict an exact count.
How it implements the components¶
A Weighted Network Propagation Model fills the discrete-graph dynamics side of the archetype:
topology_and_interface_model— it represents the medium as an explicit weighted graph of nodes and edges, the discrete connectivity structure spread must traverse.preferential_pathway_identification— running the cascade surfaces the high-weight chains and connector nodes that carry most of the spread.attenuation_and_amplification_budget— each hop's gain or loss is set by edge weight and node state, so the model tracks where transmission amplifies into a cascade versus damps out.
It runs dynamics but does not choose the intervention: it does not set the routing_intervention_policy that acts on the corridors — that decision belongs to Least-Resistance Path Simulation, whose continuous cost-surface corridor-finding is the counterpart to this model's discrete-graph cascade.
Related¶
- Instantiates: Heterogeneous Medium Propagation Routing — it simulates how an effect spreads across a discrete, unevenly-weighted network.
- Consumes: Social Tie-Strength Routing Map — a tie-strength map supplies the weighted graph this model runs its cascade over.
- Sibling mechanisms: Social Tie-Strength Routing Map · Least-Resistance Path Simulation · Finite-Element or Cellular-Automaton Model · Interface Condition Checklist
Editorial Notes¶
Form Classification¶
Form family: Analysis, Modeling & Optimization
Rationale: Weighted Network Propagation Model operates as an analytical, modeling, inference, comparison, or optimization procedure that derives insight or a solution because it a graph model where nodes and ties have different propagation weights or receptivity values.
Independent corroboration: The frozen evidence defines Weighted Network Propagation Model as 'A graph model where nodes and ties have different propagation weights or receptivity values', so its operative form is Analysis, Modeling & Optimization.
Nearest alternative: Representation, Specification & Plan — Weighted Network Propagation Model includes features of a static representation, map, specification, schema, or prospective plan that externalizes information, but its defining operation is an analytical, modeling, inference, comparison, or optimization procedure that derives insight or a solution.
Review outcome: Independent reviewer agreement; medium confidence.
Origin Attribution¶
Primary origin: Data Science & Analytics
Origin pattern: Single lineage
Present-day reach: Universal
Rationale: Vanunu et al., Associating Genes and Protein Complexes with Disease via Network Propagation documents that network science propagates weighted evidence through graph topology to score connected nodes. This is direct, mechanism-specific evidence for data science as the best-evidenced historical home of the operation—A graph model where nodes and ties have different propagation weights or receptivity values.—rather than evidence merely that the operation is useful there. The retained alternates record genuine adjacent lineages; later portability is represented separately by domain_reach=universal.
Related originating lineages:
- Computer Science & Software Engineering — Computer science and software-engineering practice supplies a parallel or contributing lineage for the mechanism's defining operation: a graph model where nodes and ties have different propagation weights or receptivity values.
- Mathematics — Mathematics supplies a historically relevant adjacent lineage or formative practice for the operation—A graph model where nodes and ties have different propagation weights or receptivity values.—but the adjudicated evidence more directly locates the defining lineage in data science.
- Operations Research — Operations research's allocation, scheduling, optimization, and decision-analysis tradition contributes a separate formative lineage to the mechanism's weighted network propagation model logic.
- Statistics & Experimental Design — Statistics, experimental design, and measurement theory supplies a parallel or contributing lineage for the mechanism's defining operation: a graph model where nodes and ties have different propagation weights or receptivity values.
- Systems Thinking & Cybernetics — Systems thinking, feedback control, and cybernetics supplies a parallel or contributing lineage for the mechanism's defining operation: a graph model where nodes and ties have different propagation weights or receptivity values.
Review resolution: The blind reviewers disagree on primary lineage (mathematics versus data_science). The defining operation is: A graph model where nodes and ties have different propagation weights or receptivity values. The researched Vanunu et al., Associating Genes and Protein Complexes with Disease via Network Propagation establishes that network science propagates weighted evidence through graph topology to score connected nodes. That source therefore supports data science as the historical origin. mathematics remains in the uncapped alternates where it contributes a formative practice, but application or governance is not itself proof of origin. origin_mode=single_lineage records lineage construction; domain_reach=universal separately records later applicability.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Researched adjudication after independent review; high confidence.
Sources consulted:
Notes¶
[n1] In heterogeneous transmission, a minority of highly-connected or highly-infectious nodes — superspreaders — account for a disproportionate share of total spread (often summarized as the "20/80" rule). It is why a weighted network model's leverage lies in the few high-weight connectors rather than in average nodes. ↩