The future of data visualisationData changes how we draw conclusions about the world, data visualisation helps us better understand data.
Flow BohlBlockedUnblockFollowFollowingOct 6, 2018Data visualization — as I’ve observed — has been changed due to developments in three broad areas.
Streaming instead of static data, changing context and better data processing and design tools.
This will profoundly change how data visualization will influence our future lives.
New contextPandora was the first human made by the gods in Greek Mythology.
As a concept this is important because Pandora serves as a metaphor of who we aspire to be.
She is a representation of ourselves, she is how we would like to be.
Very much like digital avatars in social media of the presence.
On Facebook we don’t share our worst moments or weaknesses, we show off our achievements and upload only our best pictures.
We create better versions of ourselves, like digital Pandoras.
Data visualization of related topics and users on TwitterI went to Google and other social media platforms to download all personal data about myself I could find.
1.
2GB downloaded data later, I can confidently say it’s eerily accurate but also somewhat incomplete.
Although, It’s more data about me than I expected to find.
In future this will change.
Currently, we’re creating 28 zerrabytes (or a trillion gigabytes) of data per year, at an exponential rate, therefore its easy to predict, that there will be more data available about me, in fact on all of us, creating a more complete picture of who we are.
Not only that but also who and where we are at any given time in the past.
My digital alter ego or “avatar” one day may not be as pretty as Pandora, but more accurate than any of us might wish.
Not only data on people will be more readily available but data on absolutely anything, like our cars and of course our homes.
All devices interconnected through data streams will help us improve safety, drive operational efficiencies and better understand consumer demand.
Ever growing data avalanchesDue to the staggering volume of data we’re collecting, and the sensors we use to collect it, stream-based APIs are becoming much more prevalent.
Data visualization for a conceptual car dashboard of the futureOnly a very small proportion of data we collect should be in a database.
Most of it is collected by systems that emit events, like the battery level in an electric car.
This data starts off life as a stream, and is efficiently delivered to those who might want to use it.
Putting a database in between the sensor and the consumer will become an unnecessary luxury.
Hoarding data without a clear purpose comes at a high cost.
Using data streams changes how we draw conclusions about the world.
This means that every new data point that arrives, affects our understanding of the world immediately.
Charts are plugged in to live data streams.
News ripple through markets ever faster.
Less time is used to read each article and more automation is used to generate data and even news items.
Data visualization of TensorFlow experiment called T-SNE by GoogleAll you need to communicate is a sentiment for an algorithm to pick up your story and classify it.
For example when Musk announced on Twitter that he was considering taking Tesla private, the reaction in the market was extremely quick.
Yet, picking up a sentiment in unstructured data like charts and animated videos is almost impossible for algorithms.
Charts and animations were always created for human consumption.
Not anymore.
Bloomberg has created a system called Scatteract.
It can read each pixel along an axis to reveal the information that generated each dot.
Data visualization reversed: From chart to data with Scatteract by BloombergThe system uses optical character recognition (OCR) and deep learning techniques to take numerical data points from the image of a chart and then converts the results into tables.
It’s the first time ever that data visualization is reversed.
Instead of going from data to visual now we can go from visuals back to data again.
Even though Scatteract is crude yet, it’s only a question of time for these tools to be improved and in future data visualization will be fully legible to algorithms.
Tools to shape the futureIn many product announcements from Google, Apple and BMW, more and more data will be overlaid in our physical environments through augmented reality or projection.
That means not only will data be visualized more, but the visual reality around us will be turned into data.
Data visualization of a new AR wind shield by AppleBy 2020, 70% of all the new cars will have some form of AR technology onboard, like this Apple windshield.
Smart city traffic management, like “DiDi Smart Transportation Brain” will allow for traffic participants and way finding software to make better informed decisions.
Didi Smart Traffic Lights have been installed in Chinese cities this year saving up to a quarter of waiting time.
The data of all 450 million mobile users of Didi is shared.
Including data from their connected cars and the traffic management system to create the best picture of traffic flows.
Data visualization of traffic by Didi Smart Transportation BrainUrban traffic problems aren’t just giving a picture of the presence.
Smart traffic systems allow to forecast where traffic jams will be within the next few hours.
It allows to forecast taxi and public transport shortages and can redirect other traffic participants to most effectively handle upcoming congestions.
The smarter a city becomes, the more its data needs to be analyzed and “read”.
Anomalies, patterns and spikes in data will be easily detected with data visualization.
As data availability increases so are user expectations about its usefulness.
If new solutions like Didi tells us anything about new tools, it’s that people will have higher expectations to data availability, its accuracy, usefulness and how it’s visualized in future.
The trend is clearly for new tools to provide more data to its users, not less.
They will be easier to use and therefore widen in reach, as we’ve seen with the rising interest in Tableau and Power BI.
Power BI compared to Tableau in Google TrendsIn future more than ever, professionals will be asked to make their decisions based on data, like journalists.
In times of “fake news”, journalistic storytelling becomes increasingly undermined with data.
One good example is ‘America’s Retail Apocalypse’.
Data visualization of America’s Retail apocalypse by BloombergIt paints a very clear picture about store closures around the US.
Not only that, the data alone wouldn’t be as eerie.
Painted in red and black, it’s emotional appeal is massively emphasized.
The store closures come across as threatening.
Data presented with the purpose of generating a larger appeal to a wider audience, allows for wider data scrutiny through more diverse interpretations.
Data visualization is different from pure rational analysis in that it aims at telling a story to create an emotional response.
You can have rational evidence right in front of you, but if you can’t imagine something that has never existed before, it’s impossible to stimulate far reaching action.
Algorithms can predict, humans can imagine.
Data visualization inspires imagination best.
Tools of the future will allow for more ways to slice and dice data.
This is important because with a very limited set of tools, it’s easy to fall into the trap of valuing what can be measured, instead of measuring what’s valuable.
This will allow for more accurate depictions of the world around us avoiding a Pandora style fake future of smoke and mirrors.
Here’s the talk from BigData Moscow in full:.