Function: computeNormalizedRelevanceScores()
ts
function computeNormalizedRelevanceScores(candidates, queryText): ReadonlyMap<string, number>;Defined in: postgres/src/stores/internal/compute-normalized-relevance.ts:38
Computes normalized BM25 + fuzzy/prefix relevance scores $R(L, Q) \in [0, 1]$ across candidates using MiniSearch.
Indexes directive and partition with { prefix: true, fuzzy: 0.2 }, returning 1.0 for all candidates when queryText is empty, score / maxScore when maxScore > 0, and 0 when a candidate does not match (SPEC.md §3.4.3).
Parameters
| Parameter | Type | Description |
|---|---|---|
candidates | readonly RelevanceCandidateParams[] | Candidate lessons to score. |
queryText | string | Search query text $Q$. |
Returns
ReadonlyMap<string, number>
Map of candidate id to normalized relevance score $R(L, Q) \in [0, 1]$.