Crisis of entity resolution (ER) in heterogeneous AI environments — data, financial & semantic dimensions. Part of the "entity turn" in AI: can systems reliably identify the same entity across sources, time, languages & contexts, when shared names, strings,...
More
Crisis of entity resolution (ER) in heterogeneous AI environments — data, financial & semantic dimensions. Part of the "entity turn" in AI: can systems reliably identify the same entity across sources, time, languages & contexts, when shared names, strings, or embeddings alone don't establish reference?
Synthesizes research on entity linking, entity matching, temporal knowledge graphs (KGs), retrieval-augmented generation (RAG) & multi-agent LLM systems. Identifies four anomalies: identity collapse, semantic fragmentation, temporal discontinuity & semantic vulnerability — failures that propagate through RAG/agentic pipelines from indexing to generation.
Proposes Digital Identity Optimization (DIO) & Ontology of Digital Identity (ODI) — a coherence framework (entity, representation, interpretation, trust, relation, reconstruction) as a testable integrity hypothesis, not a ready-made solution.
https://danielberanek.cz/digital-identity-optimization-dio/
Less
