Entity Resolution: From Exact Matching to AI-Powered Deduplication Webinar

Why This Topic Matters Government agencies at every level — law enforcement, courts, corrections, social services, and public health — depend on accurately linking records about individuals across databases that were never designed to talk to each other. Whether it’s matching an arrest record to a prior conviction, reconciling subject records across jurisdictions, or connecting benefits enrollment to health outcomes, the quality of that linkage directly affects operational decisions, civil rights, and public safety outcomes. Intended Audience This webinar is designed for justice and public safety information technology professionals, data stewards, analysts, and policy staff who work with multi-source data, manage identity systems, or are evaluating tools for deduplication and entity resolution. Value Promise By the end of this session, attendees will understand: The difference between exact matching, approximate/partial string matching, and probabilistic record linkage — and when each approach is appropriate How probabilistic matching (Fellegi-Sunter and related methods) works conceptually, what “blocking” means in practice, etc. How large language model LLM-based tools like LinkTransformer extend traditional methods to handle noisy, multilingual, and semantically complex data with as few as four lines of code Practical considerations for adopting or piloting modern record linkage tools in a government or justice agency context, including reproducibility, auditability, and model customization