The activity is Standardization. Standardization converts semantically equivalent values into a consistent representation so they can be compared, matched, integrated, and interpreted reliably.
In the scenario, Street, St., and Str may represent the same address concept but differ syntactically. Converting them to an approved standard reduces superficial variation before entity matching.
Standardization is commonly used for names, addresses, telephone numbers, units of measure, dates, product descriptions, geographic references, abbreviations, and coded values. It frequently precedes matching and de-duplication because entity-resolution algorithms perform more effectively when predictable representational differences have already been normalized.
Reference Data and Metadata Management are important supporting disciplines. Reference Data can define approved abbreviations or formats, while metadata documents transformation rules and intended meaning.
Standardization does not prove that the resulting address is factually accurate. It improves consistency and comparability. An address can be perfectly standardized yet belong to the wrong customer.
Therefore, standardization should be combined with validation, verification, matching, and source-improvement controls where appropriate.
Reference Topics: DAMA-DMBOK2 Chapter 13 — Standardization; Cleansing; Matching; Consistency; Reference Data Management.
===============
