Through a Google Chrome extension, users can instantly generate
summaries identifying the key insights and trends conveyed in the
graphics, in simple, every day human language. Technology like this has
the potential to vastly reduce the hours spent by people poring over
charts and tables and is another step toward putting Big Data analytics
in the hands of the masses.
The backbone of Narrative Science’s NLG platform – it’s Quill system – was initially built to generate sports and financial reports mimicking human-produced journalism. However, in a short period of time it has begun to evolve into a full NLG for BI solution.
CEO Stuart Frankell told me “More often than not the first thing you see [when opening a report] is a chart or a graph. You sit there and think ‘ok, now I have to do a bunch of work here – I have to figure out what this picture is trying to tell me’ – and I think that is a terrible user experience and a waste of time.”
“One of the things that analysts and data scientists do when they explore data through Tableau and build charts, is they explain those charts and graphs – because they don’t explain themselves. They’re communicating to people beyond just a small group of analysts - they’re communicating to their boss, or a partner, or a customer. Automated narrative generation does that for you – it teases out what’s important and interesting in your data, and actually writes it up for you. All you have to do is read.”
Automated narratives are likely to play a key part in one of the major technology trends in business today – the blurring of the interfaces between humans and computers brought about by the Internet of Things and Industry 4.0. Given the complexity of teaching humans to speak to computers in their own language computer (i.e coding), teaching computers to speak and act like us seems likely to be more efficient at breaking down those barriers of communication. Although the job of teaching one computer to talk and write like a human may be hugely complex, once it’s been done, it can be replicated infinitely and very quickly. Humans, on the other hand, have to be taught individually.
In particular, the technology is likely to be useful when it comes to alerting humans to situations which require their attention – for example when something breaks and can’t be automatically fixed. Computers and machinery will be able to explain what’s gone wrong in understandable terms, rather than the unfathomable error messages we’ve become accustomed to. Smart or autonomous cars and vehicles are prime examples of machines which will benefit from more streamlined channels of communication between device and user.
Analysts at Gartner referred to this blurring of the lines between humans and machines as creating “transparently immersive experiences” when they highlighted these concepts as key to their 2016 Hype Cycle recently - “The combination of NLG with automated pattern detection and self-service data preparation has the potential to drive the user experience of next-generation smart data discovery platforms, and expand the benefits of advanced analytics to a wider audience of business users and citizen data scientists”.
“What we really believe, and the market is starting to speak this”, Frankell tells me, “Is that narrative really is the missing piece. We believe that narrative products can dramatically enhance the user experience.”
It also ties in nicely with the fashionable idea of the “citizen data scientist” – the concept that analytics is too important to remain solely the domain of lab-bound data scientists and should be put into the hands of as many people as possible, in real-world situations, where creative thinking is likely to emerge. Data after all is useful to everyone – not merely the tiny minority who have spent years learning how to decipher it.
When it comes to working with data, the more legwork that technologies like NLG can do for us, the more time it frees up for us to do things only humans can do, such as having new ideas and implementing “big picture” strategies. It will be very interesting to see how partnerships such as those between Narrative Science and Tableau will impact how we work with data.
The backbone of Narrative Science’s NLG platform – it’s Quill system – was initially built to generate sports and financial reports mimicking human-produced journalism. However, in a short period of time it has begun to evolve into a full NLG for BI solution.
CEO Stuart Frankell told me “More often than not the first thing you see [when opening a report] is a chart or a graph. You sit there and think ‘ok, now I have to do a bunch of work here – I have to figure out what this picture is trying to tell me’ – and I think that is a terrible user experience and a waste of time.”
“One of the things that analysts and data scientists do when they explore data through Tableau and build charts, is they explain those charts and graphs – because they don’t explain themselves. They’re communicating to people beyond just a small group of analysts - they’re communicating to their boss, or a partner, or a customer. Automated narrative generation does that for you – it teases out what’s important and interesting in your data, and actually writes it up for you. All you have to do is read.”
Automated narratives are likely to play a key part in one of the major technology trends in business today – the blurring of the interfaces between humans and computers brought about by the Internet of Things and Industry 4.0. Given the complexity of teaching humans to speak to computers in their own language computer (i.e coding), teaching computers to speak and act like us seems likely to be more efficient at breaking down those barriers of communication. Although the job of teaching one computer to talk and write like a human may be hugely complex, once it’s been done, it can be replicated infinitely and very quickly. Humans, on the other hand, have to be taught individually.
In particular, the technology is likely to be useful when it comes to alerting humans to situations which require their attention – for example when something breaks and can’t be automatically fixed. Computers and machinery will be able to explain what’s gone wrong in understandable terms, rather than the unfathomable error messages we’ve become accustomed to. Smart or autonomous cars and vehicles are prime examples of machines which will benefit from more streamlined channels of communication between device and user.
Analysts at Gartner referred to this blurring of the lines between humans and machines as creating “transparently immersive experiences” when they highlighted these concepts as key to their 2016 Hype Cycle recently - “The combination of NLG with automated pattern detection and self-service data preparation has the potential to drive the user experience of next-generation smart data discovery platforms, and expand the benefits of advanced analytics to a wider audience of business users and citizen data scientists”.
“What we really believe, and the market is starting to speak this”, Frankell tells me, “Is that narrative really is the missing piece. We believe that narrative products can dramatically enhance the user experience.”
It also ties in nicely with the fashionable idea of the “citizen data scientist” – the concept that analytics is too important to remain solely the domain of lab-bound data scientists and should be put into the hands of as many people as possible, in real-world situations, where creative thinking is likely to emerge. Data after all is useful to everyone – not merely the tiny minority who have spent years learning how to decipher it.
When it comes to working with data, the more legwork that technologies like NLG can do for us, the more time it frees up for us to do things only humans can do, such as having new ideas and implementing “big picture” strategies. It will be very interesting to see how partnerships such as those between Narrative Science and Tableau will impact how we work with data.
No comments:
Post a Comment