Identify clusters of entries that are topically related but not yet linked — "structural gaps" — and surface them as suggested connections or research prompts.
After community detection, find pairs of clusters with high intra-cluster density but low inter-cluster connectivity
Highlight gap edges (potential missing links) as dashed lines on the graph
Show a "Gaps" panel listing cluster pairs with suggested bridge entries
Optionally use AI to propose why two clusters might connect and suggest new entries or links
Complements the existing "wanted pages" feature by finding missing relationships, not just missing pagesInspired by InfraNodus (Paranyushkin, WWW'19) which detects structural gaps between discourse communities to surface blind spots and prompt new insight.