What is Data Federation?

We have been hearing a lot about data science, data lake, etc…

But did you ever heard “Data Federation”?

Data Federation technology, also called data virtualization technology or data federation services, is a process composed of techniques and software that provides an organization with the ability to collect data from disparate sources and aggregate it into a virtual repository , where it can be used for business intelligence (BI) or other analysis.

The Data Federation process provides organizational capability by aggregating data from different sources into a single database so that it can be used to extract business intelligence (BI) or other analysis, adding predictive capability, pattern identification and getting answers in timely time for tactical and strategic actions in the face of market challenges.

The virtual and unifying database of information from different sources does not contain the data itself. Instead, it contains information about the actual data and its location, called metadata. The real data is left in the place of origin or in another more suitable repository.

This approach is especially useful if some of an organization’s data is stored offsite by a third-party cloud service provider or any storage service with heterogeneous technologies. It allows data aggregation analysis activity to be done quickly, without having to request synchronization logic or copy data until absolutely necessary.

The great difference in data federation and the added value that this model brings, as it allows the establishment of a centralized “library”, governed by patterns and unique semantic terms, which also establishes the unification of business concepts and definitions.

So-called Data Virtualization, which is the act of making a single call to multiple data sources and then integrating and organizing the data into a middleware layer, is also called enterprise information integration (EII). Its implementation depends on the information maturity level and the way the company uses its data.

Data Federation is one of the important matters of organizational data architecture and its argument is reinforced by the great gains that can be obtained by having a single repository, a product of the unification and standardization of segregated repositories, making the business analysis more efficient.


Source: What is a Data Federation?