Add Street Network Centrality scoring (PageRank for streets)

[Poseidon-t]

What

Score streets not just on their own walkability but on their importance to the overall safe pedestrian network — using graph centrality algorithms (betweenness + eigenvector centrality) on the OSM street graph.

Why

Not all streets are equal. A moderately safe street that connects two major safe corridors is far more valuable to fix than a very safe dead-end street. This is the same logic as Google's PageRank — a page linked by important pages is itself important.

For cities with limited budgets, this answers the critical question: "Which streets should we fix first for maximum network impact?" A street with high betweenness centrality and low safety is the highest-priority intervention — it's a bottleneck in the safe pedestrian network.

How

Betweenness Centrality:

• How many shortest pedestrian paths pass through this street segment?
• High betweenness = critical corridor. If this street is unsafe, it blocks safe routes across the neighborhood.

Eigenvector Centrality (PageRank):

• Is this street connected to other highly-connected safe streets?
• A street linking two well-connected safe zones is more valuable than one connecting to dead ends.

Combined Priority Score:

• priority = centrality_score × (10 - safety_score)
• High centrality + low safety = highest priority intervention
• High centrality + high safety = critical asset to protect
• Low centrality + low safety = local problem, lower urgency

Display

• New "Network Importance" indicator on each analyzed street
• Priority matrix: 2×2 grid of Centrality vs Safety
• Narrative: "This street is a critical pedestrian corridor — improving it would benefit the entire neighborhood's connectivity"

Technical approach

• Build weighted graph from OSM way data (already fetched)
• Edge weights: inverse of walkability score (safer = lower cost = preferred route)
• Run betweenness centrality using a BFS/Dijkstra approach
• Eigenvector centrality via power iteration (converges fast on street networks)
• These are standard graph algorithms — libraries like graphology (JS) can handle this

Relevant files

• src/services/overpass.ts — OSM street network data with way geometries
• src/utils/metrics.ts — safety scores per segment
• src/components/streetcheck/StreetNetworkPanel.tsx — network visualization
• New: src/utils/networkCentrality.ts

Complexity

Medium — graph algorithms are well-understood and libraries exist. The challenge is building the graph correctly from OSM way data and making the results interpretable to non-technical users.

Theoretical basis

Graph theory — betweenness centrality (Freeman, 1977) and eigenvector centrality (Bonacich, 1987). Applied to urban networks by researchers like Sergio Porta and others in space syntax analysis.