Principle-Based Entity Resolution in Senzing Software

This technical whitepaper provides an overview of Senzing software's principle-based approach to entity resolution, which eliminates the need for pre-training, tuning, or expert configuration. The document explains how this method can be applied across various entity types, including people, organizations, vessels, and vehicles. It details the principles that guide entity resolution, distinguishing them from traditional rules. The software operates based on expected behaviors of entity attributes, such as names and identifiers, and adapts in real-time to exceptions in data. The paper outlines the behaviors assigned to entity attributes, including frequency, exclusivity, and stability, and describes how these principles facilitate efficient and accurate entity resolution. Additionally, the document discusses the versatility of principles across different entity types and emphasizes the continuous improvement of Senzing's principles to enhance accuracy over time. Overall, it serves as a comprehensive guide for technical evaluators and architects interested in understanding Senzing's unique capabilities.