Visualising data so that businesses have something tangible to deal with is an important part of the big data/data analytics puzzle.
The importance of this is highlighted by the fact that all business intelligence (BI) and analytics vendors offer data visualisation tools to some degree or another.
Users can pick from SAS, SAP, AWS, Teradata, Tableau or numerous other vendors and they will find either a suite of data visualisations included or the product tailored to specifically delivering visuals.
Put simply, data visualisation is the presentation of data in a pictorial or graphical format. It has been around for centuries from maps and graphs in the 17th Century to the invention of the pie chart in the early 1800s.
Technology has really lit the touch paper of what is possible with data visualisation. Recently for example the people at the Ordnance Survey produced a map of Mars at the scale of one to four million.
This was achieved by using NASA open data to create a one-off paper and digital map of the Martian landscape. The purpose of this was to see whether the OS style of mapping has potential for future Mars missions.
Cartographic Designer, Chris Wesson, spent a couple of months designing the maps, he told CBR that it was built from a combination of data but elevation data formed the bulk of it.
Wesson said: "There were a few challenges, not huge. Understanding the data was a challenge because on Mars it is hard to find reference points for elevation data because there are no oceans."
Cartography is one of the most traditional forms of representing data in a format that the masses can understand and that is, at its core, what data visualisation is doing.
Companies like Tableau have experienced great success as the big data market has grown, mainly because the company is making it easy for the average user without any technological background to be able to get to grips with valuable data.
It serves an important role in the democratisation of data and in actually bringing value to data. The problem is that data on its own holds no real value, analytics and processes need to be applied to it to remove it from essentially being an expensive paper weight.
According to SAS, data visualisation is important because of the way the human brain processes information, visualising large amounts of complex data is easier than poring over spreadsheets or reports.
In data that could include thousands or even millions of variables – a picture is worth a thousand words.
There is typically considered to be four Vs when it comes to big data; Volume, Velocity, Variety, Veracity. However, another V could simply be added to that – Visualisation.
Our attention spans are shortening (around 8 minutes) alondside the explosion of data and the need to understand it, so visualisation makes perfect sense because we essentially can’t comprehend the size and scale of most data.
For example, did you know that data will grow to 44 zettabytes by 2020 from just 4.4 zettabytes in 2014. One zettabyte is equivalent to a billion terabytes, which is equal to a trillion gigabytes – have you got your head around that?
Well here is a data visualisation to help you understand it.
This can be something simple but full of value but that doesn’t mean that data visualisations instantly equate to value or that it is done correctly – there is an art to it.
Anthony Deighton, CTO of Qlik told CBR in an interview last year: "Most people do a crappy job at data visualisation."
"Most people are not experts in data visualisation. Microsoft Excel is a disaster when it comes to this stuff. Most people’s visualisations on Excel are useless."
Qlik provides data visualisation tools for its product that enable enterprises to automatically create graphics from their data, and that data could be on customer insights, information about employees or from numerous other areas.
One of the criticisms of data visualisation tools is that while they produce pretty pictures they don’t really influence the decision making process or force a decision.
SAP is one company that has been vocal in the past on this subject and it has been working to integrate visualisation and analytics tools into S4 HANA so that it solves that exact problem.
A checklist for data visualisation preparation should be in place to help inform how the data is framed.
Firstly ask whether there is an understanding of the data that is being visualised, this includes knowing its size and cardinality. Ask who the audience is and how they process visual information – basically will a pie chart or bar chart suit or maybe it is better to go with a moving visualisation.
What type of information of information you are presenting should be asked. Is the user trying to see relationships among data points, to compare a set of values or to track rises and falls over time?
Data visualisation shouldn’t be underestimated and if done correctly can help deliver true value from big data.
