In Memory Entity Resolver
Entity resolver that remembers entities it's been asked to resolve and tries to reuse them. Useful for deduplicating entities within a single session.
Uses name matching including:
Label compatibility checking
Case-insensitive exact matching
Name normalization (removing titles/suffixes)
Partial name matching (e.g., "Holmes" matches "Sherlock Holmes")
Levenshtein distance for fuzzy matching
Not thread-safe: it keeps a plain mutable map of remembered entities, so a single instance must not be shared across concurrent callers. This is by design — PropositionPipeline.process builds a fresh resolver per run and only calls it from the serial resolution stage, so the memory stays scoped to one run and never sees concurrent access.
Parameters
Configuration for matching thresholds
Types
Functions
For each of the given entities, resolve them to either an existing entity or a new entity.