There is a dramatic difference between analysing and understanding data

Current data visualisation tools have progressed past X/Y charts, bar graphs, scorecards, and dashboards and the terms machine learning (ML), deep learning, artificial intelligence (AI) are being used to describe the rise of the machines to uncover multi-dimensional insights even the best data scientists cannot easily see.

So how do you make sense of the complex, mind boggling, often siloed, mountains of multi-dimensional data? The answer is visualisation using statistical analysis as the foundation, to formulate a predictive perspective and to discover correlations between disparate data.

iTWire asked Julian Quinn, Vice President of visual analytics and business intelligence company Qlik about how to reveal the insights – the whole story that lives within data.

Qlik QuinnQuinn is qualified to do so having nearly 40 years in IT working in high levels positions for Unisys, Novell, Netscape, BEA, Borland, Adobe, Veeam, and now as Regional Vice President for Asia Pacific of Qlik based in Singapore.

Q. Lately, everyone is talking about AI and machine learning ML and the fear that the rise of the machines – AI-driven robots – will take over the world? When will it happen? Should we be alarmed?

The threat posed by robots is more alarmist than accurate. However, they will eventually take over boring, repetitive and unsafe job – jobs that humans don’t really want to do requiring those affected to undergo reskilling. But this is nothing new. The reality is that machines have been transforming the nature of work since the industrial revolution.

Like the previous “revolutions” – mechanisation, mass production, automation and now robotization (with supporting AI/ML), the introduction of machines to our work will simplify tasks and processes, and help us to become more efficient.

If you are doing manual or factory work you would be a little apprehensive of such change. But the other revolutions have gone OK and we should look to understand it and come to terms with how automation can assist us in our jobs. Aren’t you happy you have to have a dishwasher or a robotic vacuum cleaner? If we get this right, robots and AI will help us progress further and faster than ever before.

The AI/ML conversation started years ago but has come into the spotlight once again as they become both recognised as critical tools to leveraging the masses of data we create and collect, especially complex multi-dimensional data, that even the brightest humans simply can’t grasp as well.

But I prefer to use the term augmented human intelligence instead of artificial intelligence. Humans will still be the ones to truly understand the data.

Q. Are you saying that the rise of the machines needs a corresponding rise of the people to understand data – data literacy?

Data literacy is “the ability to read, work with, analyse and argue with data” according to Rahul Bhargava (MIT) and Catherine D’Ignazio (Emerson).

In short data without analysis is just data – you must derive meaningful information from it to be of any value. Data is being generated and collected at an exponential rate. This implies that we all need some ability to sift through it, to uncover insights, and derive value. In the post-fact/post- truth era analytics has never been more important.

So, in the data and analytics space, automation will help close the gap between the vast amounts of data created and our ability to consume [understand] it.

But that is not enough. People need to become more data literate. Although data literacy in Australia is more advanced that many other countries, there is always room for improvement. Given that Australia is a gateway to Asia Pacific for numerous global corporations, it is vital that our workforce keeps up with, and understand the changing landscape of data.

Q. How does visualisation help data literacy?

There is a dramatic difference between analysing and understanding data, and it is problematic to assume that people alone can completely comprehend the level of data presented to them. Visualisation helps democratise data.

Realistically, it is nearly impossible for a human to be able to provide insights across multiple data sources, to make sense of data with statistical analysis, to formulate a predictive perspective or to discover correlations between entirely separate business units.

This is where data literacy and automation become an important asset. Machines and robots can derive hidden perspectives from the data, which provides the foundations people to arrive at different ideas and answers.

In practice, AI/ML would find the links and present them is a way that is meaningful to humans. That is not a graph or plot but a visual analytics platform that leverages data from multiple sources and presents it in a way we can grasp the significance.

This does not mean exclusively using machines to fill the data literacy gap. The crux is leveraging them to empower and connect all the people, all the data and all the ideas to truly gain actionable insights from silos of data, as well as deliver solutions that are relevant to specific businesses and scenarios in a timely manner – putting the data in context.

A visual analytics platform that is associative and dynamic in nature ensures that everyone has complete flexibility to fully explore all the possible insights across data and data sources.

Q. Are you suggesting that everyone needs to become data literate, especially after machines take over?

Absolutely. Your next big data-driven idea will come from someone at the edge of your organization who sees things differently based different experiences, biases, and view of life. It’s incumbent on you as a leader to capture those ideas. Shine a light on data literate people, and encourage more people to become data activists.

To achieve the productivity we need, effectivity and efficiencies that insights derived from data literacy can provide, employers need to evaluate their workplace and grasp how they can enable their team.

As opposed to the common tactic of addressing and implementing purely from a top-down order, data literacy projects need to come both ways – employee buy-in and a bottom-up approach in addition to a top-down one.

This is where education, collaboration, and sharing play a key role in demonstrating to employees the impact of data literacy in their context – on the company as well as on their specific roles. In effect, employees need to understand what they can do to make a difference.

Given the advanced digital technology available today, lack of data illiteracy can no longer be an excuse for employers. But we know technology isn’t enough. It’s incumbent on our leaders to build a data-driven culture that connects data with people to capture their ideas.