The mapping statistics, in full.

Every correspondence carries a statistical profile: how often the English morpheme is realised as this Vietnamese one (EN→VI fit), how dedicated that Vietnamese morpheme is to it (VI→EN fit) — two equivalent directions of the same bidirectional mapping — their harmonic mean, association controlled for base rates (NPMI), and a sample-size-aware floor (Wilson lower bound). The composite fit index balances all of these and is damped — never zeroed — for analyst-only evidence. Sort any column; filter by band.

English morpheme nodes — volatility & expectability

Vietnamese Han-Viet nodes — promiscuity & specificity

All metrics are recomputed from the observation ledger after every committed batch, so nothing is a black box. Probabilities are reported raw; the Wilson lower bound carries the uncertainty rather than hiding it behind smoothing — the formal expression of “ping more before trusting a thin edge”. Drawn live from the English–Vietnamese Analytical Mapping database. See also the explorer, the bridge, and the ripple.