Populating Databases Automatically with Realistic Data Using Metadata-Driven Methods

This is the authors' abstract. We don't add key points for this paper.

This research investigated the feasibility and utility of a metadata-driven “fake data“ generator. The researchers created a prototype for a data generator whose job is to be pointed to a database and then generate fake but appropriate and realistic data. Using metadata analysis, the generator was able to make autonomous and smart decisions as to how to populate the various tables and fields it encounters, populating them with realistic data. This promises great utility to programmers, QA testers, and database administrators for various use cases such as testing, debugging, database design validation, load testing, and performance scaling.