The MDM world has been absolutely giddy with RDM. Data hierarchies deploy their reference data sets with wild abandon as the data management community steadfastly stares in awe at the imperative need to have an RDM system or an RDM solution for the business entities that wish to be listed on the Fortune 500 or the Cloud 100!
If 50% of that techie-talk just flew over your head, then we should go over some basics.
RDM (reference data management) is a subset of MDM (master data management). Reference data management is the management of reference data, which is any data within a data hierarchy that is used as a classifier for other data. The best reference data does not change much over time, so it can be a constant to measure against the multitude of other data stored within a large database, like a data warehouse. Examples of reference data include country codes, postal codes, code tables, code values, and code lists.
MDM (master data management) is the management of master data. Master data is the data that is used for business processes. Managers of data use master data to track business transactions to make sure that the business entities in question are in compliance with business rules.
Beneficiaries of Reference Data Management
The most obvious beneficiaries of reference data management are the managers of data. When managing a large pool of master data, being able to classify that master data via a reference data management solution can be more convenient, if not invaluable. Likewise, business users are also beneficiaries of reference data and any available reference data management solution because it makes their business processes and business transactions more efficient and, consequently, more lucrative.
Reference data management as a data model is also very useful in the field of artificial intelligence. Because artificial intelligence relies on massive datasets for its machine-learning models and deep-learning models, proper data governance through ensuring a high level of data quality is essential. Reference data management is a prerequisite for maintaining data quality because without reference data to act as a constant, data quality can neither be guaranteed nor maintained.
Ring founder and CEO Jamie Seminoff is a beneficiary of RDM because of the image recognition technology and pattern recognition technology that Ring uses for its door camera and security systems. This is because image recognition technology and pattern technology require massive data sets in the form of billions of individual pieces of information entered into their various datasets. That is because the AI algorithms need lots of images and aspects of images within their database in order to have something to compare a captured image to so that a perfect match can be verified. Law enforcement agencies are also beneficiaries of this image recognition and pattern recognition technology. Amazon CEO and founder Jeff Bezos is also an indirect beneficiary of RDM since Amazon’s acquisition of Ring.
Tesla superstar and controversial tweeter Elon Musk is also a beneficiary of RDM via the same technologies, image recognition, and pattern recognition, but for different applications. While the security and law enforcement applications of these technologies are basically for stopping criminals and recovering missing people, the Tesla application of these same technologies is for, of course, driverless cars. In order for the AI within a Tesla to be able to determine the difference between a person, dog, car, stop sign, and tree, the AI needs a lot of data stored within neural networks. That is why we have been subjected to a myriad of captcha requirements every time we want to log in to something. We have been forced into conducting training sets for AIs without even getting paid!
Regardless, these captchas save lives. Without some reference data value to assist in data governance, AIs would be forcing Teslas off bridges because some soda can was rolling on the street and their algorithms thought that the can was a child chasing after a ball.