Thursday Seeds | Graphers that POV

What are graphers? 

They are typically numbers people but really relate to data best in graphic form. They understand the need to harness the data that many companies are already storing. Instead of using SQL and traditional databases, they use graph databases which structure the data in a design pattern. Using these design patterns, graphers use a different method to pull that data out for presentation so that business leaders can make specific relevant decisions about their business.

My usual go to when I refer to the need of a knowledge graph is “graphers gonna graph”. This takes it all a step further. Because not only do they graph, that was step one, they also graph with multiple POV. Graphers that POV.

What is POV? 

Point of view.  Usually it’s hero and villain. Maybe a few others, say less than 5  additional characters.

What if it were possible to track 128 facets like the shape of a cut diamond? Then we could really see what it is we are learning.  This might require the help of AI. It might actually be cool and it might actually be graphic too! Let’s take a look see.

Why do we even need all this? It’s possible that we could either leverage some business method that no one has thought of. It’s also possible that we could innovate in directions that many wouldn’t pick up intuitively looking at the data in a tabular form. Looking for the anomalies? They pop out specifically when looking at the data in graph form. Statisticians sort of all ready know this. The problem is that our level of data count is massively larger than ever before. We call it big data for a reason. One of the rules of good statistics is a decent data size. Today’s data is yesterday’s too much data to master. Thus the need to innovate. To be creative, and to connect the data in patterns not used before.

In a relational database, list of values is king.

Sure, you can create dependencies and connect things, but with pattern design, it’s intuitive to see if there are issues. The devils that lurk in the details become visible with a graphical sifting of the patterns. And while that only seems like it makes sense, if we liken it to musical patterns, it’s easier to understand.

If a person is learning an instrument, the tune they wish to play must first have a pace set.  A counting or tapping out of the beat is the beginning. Then each note must be played on the correct pitch. And that pitch that is made, must also coincide with the set key. This is done with a series of notes, that our ears, some more tuned than others, can hear and discern what the key pattern is. If the note is played on pitch but outside the key, it’s noticeable. If the note is played not on pitch but within key, if the tune is not known, it might not be detected. Or if it is, it could be heard a harmonic. The more the tunes are known, the more this kind of detection can be made. That is why the idea of looking at the same data in 128 different view points, may not be so far fetched or a bad idea. It’s not overkill, and AI won’t complain about it being monotonous either.

Next week, We can talk about data examples that might fit this model. Data patterns might be the step in understanding. Clear as mud? Stay tuned.