Data Transformation

Go to Terms and Definitions

A Data Transformation applies one or more operations to source Data to produce resulting Data having a changed form, Structure, representation, or value.

Transformation may include normalization, conversion, aggregation, filtering, enrichment, redaction, anonymization, encoding, decoding, compression, decompression, calculation, or another operation that produces distinguishable resulting Data.

Data Transformation remains distinct from Data Transfer. Transformation changes the Data. Transfer changes the location or control boundary through which Data moves. Transformation also remains distinct from interpretation: transformation produces Data, while interpretation derives meaning or conclusions from Data.

operation that produces data having a changed form, structure, representation, or value from source data

Dido Solutions, Inc. and Jackrabbit Consulting, Inc., derived for DIDO-TE and informed by ISO/IEC data-processing and metadata concepts

A Data Transformation should preserve Provenance and Traceability from resulting Data to source Data and to the transformation rule, parameters, implementation, actor, and time when those facts are required to reproduce or evaluate the result.

A reversible transformation permits recovery of the source representation from the result. An irreversible transformation does not. Reversibility does not determine whether a transformation is authorized.

DIDO-TE transforms a source event stream into normalized Test Input records. The transformation record identifies the source events, mapping rules, parameter values, resulting records, and detected exceptions.