Prototype - Data Prototyping

Data Prototyping

A data prototype is a form of functional or working prototype. The justification for its creation is usually a data migration, data integration or application implementation project and the raw materials used as input are an instance of all the relevant data which exists at the start of the project.
The objectives of data prototyping are to produce:

• A set of data cleansing and transformation rules which have been seen to produce data which is all fit for purpose.
• A dataset which is the result of those rules being applied to an instance of the relevant raw (source) data.

To achieve this, a data architect uses a graphical interface to interactively develop and execute transformation and cleansing rules using raw data. The resultant data is then evaluated and the rules refined. Beyond the obvious visual checking of the data on-screen by the data architect, the usual evaluation and validation approaches are to use Data profiling software and then to insert the resultant data into a test version of the target application and trial its use.