Identity resolution and entity resolution are two terms that are often used interchangeably, but they actually refer to two different processes. While identity resolution is the process of linking data and online activities to specific people or users, entity resolution is the same process that applies to linking disparate sources and fragments of information to specifically identified elements or entities, human or non-human. To understand the difference between identity resolution and entity resolution, let's look at an example. Imagine a police investigation in which fingerprints are found at a crime scene.
The fingerprints are “references to the people who left them”, but they are not in themselves people. Entity resolution is the process of determining if two entity references are for the same entity or for different entities. In this case, the police would use entity resolution to classify the fingerprints into two groups that refer to two different people. Identity resolution is closely linked to entity resolution, but it is more specifically related to people. Identity resolution is used to detect identity theft and fraud, as well as for customer data integration (CDI) and master data management (MDM).
It is also used by corporate service providers to resolve the names of organizations despite different representations, spelling errors, abbreviations, and typographical errors. Entity resolution is also used in large database solutions for supply chain management. It is used to consolidate supplier data into data silos spread across multiple business units, regions, geographies, and categories of parts and materials. It can also be used to reconcile products, compare their prices, and decide which vendor sells the cheapest. Finally, entity resolution is used in medical record systems to develop and maintain a master patient index (MPI). This process is not just about matching unknown identities with known identities; it involves a series of algorithms, odds and scores that are applied to find and determine any associated record. In summary, identity resolution refers to linking data and online activities to specific people or users, while entity resolution applies to linking disparate sources and fragments of information to specifically identified elements or entities.
Neo4j has an integrated library for graphic data science algorithms such as identity resolution that can be used directly in your database using the Cypher syntax.