Generally accepted synonyms for "identity resolution" include:

Identity Resolution Glossary

  • Address normalization - is the process of standardizing and transforming address data into a consistent, structured, and uniform format. Given that addresses can be represented in multiple ways with varying levels of detail, misspellings, or abbreviations, normalization ensures that they are stored and processed in a standardized manner. Address normaliztion is often carried out immediately before geocoding. 

  • Attribute - a property or field of a data record, such as FirstName, LastName, EmailAddress, DateOfBirth etc. In database terms, an attribute represents a column in a table. 

  • API - stands for Application Programming Interface and is a tool or protocol used to allow two software systems to interact with each other directly. Tilores specifically uses a GraphQL API, which is used to both ingest data and to query individual entity/identity graphs. 

  • Batch - in contrast to real-time processing, rather than processing individual records as they are received, records are processed together at the same time. The frequency between batches could be seconds, to days. A batch-based identity resolution system would be considered to always be out of date - as soon as one batch starts, that means that newly received records are awaiting processing.

  • Data Deduplication - often used as a synonym for identity or entity resolution, data deduplication usually means the identification and elimination of redundant copies of data records. In Tilores we have a particular meaning for data deduplication

  • Data Record - a single thing in a database table, for instance a single customer account. In database terms, each data record, would be represented by a single row. A data record alone, or record-linked to other data records, makes up a entity. 

  • Data silo - refers to any database or system, where data is held seperately and not connected to other systems containing the same record type. Common data silos might include sales CRMs, customer support CRMs, transaction databases, marketing databases etc. Identity resolution systems aim to overcome data silos by connecting to these silos and providing a unified "source of truth" about customer data. 

  • Data warehouse - is a centralized repository that stores data integrated from various disparate sources. It's designed to support business intelligence activities, primarily the analytical processing, reporting, and data querying. Unlike regular databases which often prioritize transactional processing, data warehouses are optimized for data analysis and query processing.

  • Edge - a connection between two records, or nodes, in an individual entity/identity graph. It is created by the triggering of one or more rules. 

  • Entity - a real-world thing as represented by at least one data record. This could be a person, a company, a product, a transaction etc.

  • Entity fusion - selecting the "true" value for an attribute such as the name, e.g. chosing "Steven John Renwick" from the following names: "Steven Renwick", "S.J. Renwick", "Steven J. Renwick", "Steven John Renwick"

  • Fuzzy Matching - is a technique used to determine how similar two strings of text are to each other. This is often used in situations where it is not possible to perform an exact match. See more various Fuzzy Matching algorithms

  • Geocoding - means the enrichment of address data with geographic coordinates. This makes matching of records based on geographical distance possible. 

  • Geomatching - when physical distance between two records is used as one of the rules for record linkage. e.g. if two records are located within 100 meters of each other (based on geographic coordinates) they might be matched. To do this, records might need to first be geocoded. 

  • Golden Customer Record - another name for a complete identity graph of a customer, including linked data from all sources. Also known as unified customer data or a customer 360 degree view. 

  • Graph - sometimes more specifically "entity graph" or "identity graph", a graph represents all the linked and deduplicated record data of one entity, together with its edges, in an identity resolution system. 

  • Identity - an entity that refers to a person is usually referred to as an identity. i.e. identity resolution is entity resolution as applied to individuals. 

  • Node - when multiple data records are connected together during identity resolution, each record may be considered a "node" in an identity graph. The entire identity graph can be found by searching for any individual node. 

  • Real-time - in identity resolution this means the ingestion and processing of individual record data as it is received. The actual speed with which it is processed does not define "real-time" (but should be as fast as possible), rather that the records are ingested as the same rate as they are received. What is NOT considered real-time is batch-based processing, no matter how frequent or fast the batches. 

  • Record Linkage - often used as a synonym for identity or entity resolution, record linkage means the identifying and linking records from one data source with another or within a single dataset that pertain to the same entity. In Tilores we have a particular meaning for record linkage.  

  • Rules - are used to create edges between data records. They may include exact or fuzzy matching based on specific record attributes. Within Tilores only one rule needs to be triggered to create an edge between two records. 

  • Temporal matching - using time as one of the rules when linking records. e.g. if two records are created within a certain timeframe (and other rules are satisfied) they could be matched. This may be especially useful for transaction monitoring. 

  • Transformation - is a fundamental step in the data integration and preparation process. The primary goal is to improve data quality, before deduplication and record linkage takes place. Key steps include: cleaning, standarization and normalization. 

  • Transitive linking - the mostly linear connecting of several nodes together, e.g. from A -> B -> C -> D and so on. This means that searching with the data of record A, would also return the results of the record D.

Weekly webinar

  • Want to see Tilores identity resolution software in action?

  • Have questions about how Tilores could work for your business?

  • Eager to meet the people behind our technology?

Tilores webinar

Identity resolution for fraud prevention, KYC and marketing.

Get the latest updates

©2023 Tilores, All right reserved.