<img height="1" width="1" style="display:none;" alt="" src="https://px.ads.linkedin.com/collect/?pid=4011258&amp;fmt=gif">

What is connected data and why is no one doing it?

We are living in a relational data world. The sad thing is, that the majority of companies have data that is more like a Graph or Network in reality. Although this might be considered a moot example, there is a good reason why Google, LinkedIn and Facebook all fundamentally base themselves off the Graph structure - it is because it is the super structure, it is the structure that supports all sub structures. The trick that these 3 big players have identified is that the Graph is at the core, but it is supplemented and supported by other types of technology that can fill in for the parts where a Graph falls down.

Connected data is the idea that we can build up a network of discrete entities that are connected via relationships. Relationships are never always direct, they can be indirect - and if the intention is to find hints to relationships that could exists, then this will never be something that is direct.

Connected Data is at the core of so many different modern use cases. However it doesn’t mean that people need connected data to solve their problem, but rather they want to project connected data into a format that works for them. A good example would be machine learning. The majority of machine learning platforms would like flat, tabular data as the input. The good news is that because the Graph is a Super Structure, it can always “Downcast” to something that is off lesser sophistication, like flat tables.


Please click below to read the full whitepaper.