Critical Requirements for Identity Resolution
By Robert Barker, Infoglide Senior Vice President & Chief Marketing Officer
I’m back from the holidays with more insights about identity resolution.
Several times in past posts we’ve compared and contrasted Identity Resolution with other market spaces that it’s sometimes mistaken for, including MDM, data integration, data warehousing, name matching, and data quality. It was worthwhile to differentiate identity resolution, but the next step is to outline the critical requirements that any identity resolution solution has to address.
Below are the requirements for a comprehensive identity resolution solution. They’ve been grouped into four categories: identity matching, relationship detection and resolution, decisioning, and business process integration.
Identity matching
1 similarity (non-exact) matching
2 scrambled data (e.g. street name and street number reversed)
3 needs no special handling of missing data
4 default weightings (e.g. street name versus street number)
5 user-configurable to address unique needs
6 deterministic analytics for multiple attribute types
7 mathematical analytics
8 multiple attribute-specific algorithms invoked as need
Relationship detection/resolution
9 leverage and capture valuable ambiguous identity information
10 deterministically show derivation of detected relationships
11 anonymous resolution
Decisioning
12 user control over decisioning
Business process integration
13 fully integrated end-to-end solution in one product
14 optional live remote access to disparate data sources
15 integrate into existing infrastructures
16 use industry-standard web services and APIs
17 separate repository not mandatory
18 alert-based exception reporting
Products in adjacent spaces like MDM and data quality often address a small subset of these requirements, but the number of integrated offerings that address all of them is a very, very short list!
Subscribe to this blog