Unified Scores table

The Unified Scores table records scoring for all pairwise comparisons and match categories for all scorable record pairs within an identity graph.

The Unified Scores table is added to databases in the Customer 360 page when:

  1. The Stitch QA database is added using the “Stitch QA” database template.

  2. A customer 360 database is added using the “Customer 360” database template.

  3. A table is added to a custom database using the Unified Scores table as a passthrough table.

Use with Stitch QA

The Unified Scores table is the foundation for the Detailed Examples table in the Stitch QA database, which is the starting point for any quality review of Stitch output and is an important part of the Stitch QA workflow.

Use with Customer 360

The Unified Scores table enables using the Amperity ID to discover information about how groups of records were scored and to verify those scores against other data points.

Important

The Unified Scores table should not be used outside of the Customer 360 page. It should not be used to build queries or segments or be configured to send data to downstream workflows.

Column reference

The Unified Scores table has the following columns:

Column Name

Data type

PII

Description

Amperity ID

String

The unique identifier assigned to clusters of customer profiles that all represent the same individual. The Amperity ID does not replace primary, foreign, or other unique customer keys, but exists alongside them within unified profiles.

Note

The Amperity ID is a universally unique identifier (UUID) that is 36 characters spread across five groups separated by hyphens: 8-4-4-4-12.

For example:

123e4567-e89b-12d3-a456-426614174000

Source1

String

Source2

String

PK1

String

PK2

String

Score

Float

A score has a value from “0.0” to “5.0” that represents the strength of the match Stitch assigns to a record pair during identity resolution. A higher score always indicates a stronger match. This holds across the entire range: a 3.0 match is one point stronger than a 2.0 match, just as a 3.1 match is one point stronger than a 2.1 match.

The record pair score correlates to the match category, which is a classifier applied by Amperity to individual record pairs. The record pair score corresponds to the classification:

  • 5 for exact matches

  • 4 for excellent matches

  • 3 for high matches

  • 2 for moderate matches

  • 1 for weak matches

  • 0 for no matches

A higher value reflects a stronger signal from the machine learning model, so a score of 4.58 is a stronger match than 4.55. When two record pairs receive the same score and cannot both merge (for example, because of a conflicting given name), Stitch resolves the tie deterministically by selecting the match with the lower alphanumeric cluster ID, so results stay stable across runs.

Note

Scores are shown for records that end up in the same cluster, including any scores that are below threshold. Scores are not shown for records that do not end up in the same cluster.

Also in: Detailed Examples

Match Category

String

A match category is applied to individual record-pair comparisons discovered by deterministic and probabilistic matching strategies during identity resolution.

Match Category

Description

Exact

Amperity has the highest confidence. Records represent the same person because all profile data exactly matches.

Excellent

Amperity has near perfect confidence that records belong to the same person, despite some profile data not matching.

High

Amperity has high confidence that records match, despite some profile data not matching.

Moderate

Amperity has moderate confidence that records match, due to weak or fuzzy matches between unique customer attributes, such as email, phone, or address.

Weak

Amperity lacks confidence, but if asked to guess, Amperity would assert these records do belong to the same individual, because they match on non-unique customer attributes, such as name, state, or ZIP code.

Non-match

Amperity has high confidence that these records do NOT match, because core profile data is in conflict.

Also in: Detailed Examples

Match Type

String

The score assigned to matched records is the match type. Possible values: “scored”, “scored_transitive”, and “trivial_duplicate”. Records assigned a “scored” value are directly connected by deterministic or probabilistic matching. Records assigned a “scored_transitive” value are transitively connected.

A match is assigned “scored_transitive” when that match was not identified during blocking. For example: three records (A, B, and C). If records A and B and records B and C were identified as matching during blocking, all three records will end up in the same group of records for pairwise comparison. Records A and C have a transitive connection.