Every day we produce an almost unfathomable amount of data. Posting on Twitter, Facebook, Instagram and YouTube. Commenting in chat rooms; blogging; trading stock tips; and decorating hacks in niche forums. We broadcast what we’re eating, feeling and doing from our GPS-equipped smartphones, sharing maps of our runs, photos from shows, and news that gets us cranky or inspired.
The details of our passing moods are all there, creating a vital if virtual public pulse.
Dr Brenton Cooper’s Data to Decisions (D2D) CRC team checks this pulse and, by extracting signals from our collective digital footprint, shows where we’re going next.
Are we gearing up to strike? Or celebrate? Is disease spreading? What effect will an interest rates hike have? Are we about to toss out the government, or move money out of the market?
Whatever the social disruption, D2D CRC’s Beat The News ™ forecasting system can issue a warning – before it happens. In March 2016, it accurately forecasted the impact of an anti-coal port protest in Newcastle, NSW. The following May, no ships could move during the protest blockade, costing an estimated $20 million.
“This warning system tells you what might happen, when it will happen and why.”
Social media monitoring is already a billion-dollar industry, and Cooper, who is D2D CRC’s Chief Technology Officer, knows “there are plenty of tools that help you understand what’s happening right now. But this tells you what might happen, when it will happen and why.”
This sort of heads-up will be invaluable. D2D CRC’s first collaborators are Australia’s defence and national security agencies, whose analysts now have a Beat The News ™ dashboard that sifts through about two billion data points a day.
“These are people paid to understand the political climate, but they can’t read everything,” explains Cooper. “That’s where machine-enablement certainly helps.”
Maybe the agencies are watching Indonesian politics and want to know if there might be some unrest in the capital Jakarta. Beat The News ™ analyses a huge volume of open-source information, combining structured and unstructured data from a wide range of sources. It geo-locates posts, extracts key words, topics, trends and hashtags, and measures sentiment.
“Once we’ve done those types of data enrichments, we then pump it through a variety of models,” says Cooper, “to automatically and accurately predict the answer.”
The potential applications are many, so the CRC recently trademarked Fivecast™ – “as in forecast, only one better,” says Cooper – to take the system to market, whether as a spin-off company, licensing to a partner, or licensing the IP to a third party.
US company Dataminr has raised more than US$130 million from investors for its real-time analytics, but Cooper says Fivecast™ will offer a further capacity – event prediction. It’s the only predictive geopolitical risk analytics platform on the market. Corporate risk consultancies are already interested. Their clients include global mall conglomerates alert to anything that might stop people enjoying their shopping.
Find out more about Beat The News ™ at d2dcrc.com.au
– Lauren Martin
You might also enjoy ‘Disrupting Terrorism and Crime’. Sanjay Mazumdar, CEO of the Data to Decisions CRC (D2D CRC), takes a look at what the national security sector can learn from Big Data disruption.