Extracting meaning from data: The lowest common denominator – Featurespace

October 26, 2016

Extracting meaning from data: The lowest common denominator

Martina King, CEO at Featurespace - a global leader in machine learning fraud prevention using adaptive behavioural analytics.
 
Society functions by applying rules to everyone. The poor behaviour of the few leads to the good being treated as bad. It's deeply frustrating. The good are all tarred by the same bad brush. 
 
The world of risk management is a perfect example. The behaviour of a few rogue organisations has led to an explosion of regulation and red tape in order to ensure that consumers are protected. While protecting consumers is good, red tape can quite often be obstructive. Red tape is especially bad for the companies that would only have taken the right decisions to support their precious customer base. 
 
The same applies in our day to day lives. Criminal behaviour such as hacking, card theft or fraud by a small minority results in sweeping rules-based systems that lead to genuine transactions being incorrectly blocked. These rules are continually layered, one on top of the other, unlikely to ever be reversed, which can cause considerable and unwarranted inconvenience to the consumer. 
 
Our lives would be much easier if we could identify this rogue behaviour, and deal with it, at the time it occurs and sort it out at the level of the rogue individual. Until now, there haven’t been any effective solutions that can address the issue on an individual level, hence we are all treated by the rules written for the bad behaviour of the few – the lowest common denominator. 
 
The confluence of data and computer sciences, however, along with the extraordinary inventions of people such as Cambridge's Head of Signal Processing and Applied Statistics, Professor Bill Fitzgerald, his PhD students and Featurespace’s founder Dave Excell, have taken us to a point where we can weed out bad behaviour by understanding the good. It's a 'the world isn't flat!' moment in extracting meaning from data. 
 
By using adaptive behavioural analytics, with machine learning to understand an individual's good behaviour, across all channels, companies and organisations can take steps to better serve and protect their customers. By quietly watching all that is good, it is possible to spot the bad and deal with it, leaving the good blissfully unaware and able to continue unperturbed.
 
As well as the considerable improvement this can bring to the customer experience, in the payments sector we have calculated that $12bn of annual business improvements can be made by using technology that identifies the anomalistic or rogue behaviour and deal with it, as it occurs. 
 
It's a positive shift and for once it is a shift that rewards the good.  

Find out more about the ARIC platform.