Big data has revolutionised fraud prevention, according to Alan Harton, VP international of software provider I4C Analytics.
Harton told delegates at Computing's annual Big Data Summit, that fraud prevention had slowly changed from being descriptive - asking "what has happened", to being prescriptive - asking "what actions can be taken to prevent fraud from taking place".
This change from descriptive, to diagnostic, to predictive and now prescriptive, requires less and less human input, Harton said, although he emphasised that human input is still a key component of any analysis.
Harton explained that when consumers wanted a new phone contract or credit card previously they would submit an application that would either be accepted or rejected using the consumers' profile history.
Now, he said, there are several different techniques that are used to check for fraudulent activity, thanks largely to the amount of data available to analyse.
Harton said all applications can be checked for typical patterns of fraudulent activity and anomalies, such as behaviour out of character with a particular customer, or spending patterns that are more irregular than those of the majority of consumers.
Applications can also be checked for "link analysis", which is where companies make links between human beings and channels. For example, if there is a call centre salesman making hundreds of sales without any invoices to show for it, this is likely to be spotted within the "link analysis" part of the screening process.
However, he warned companies that are attempting to tackle fraud to think about how much they are spending to recuperate funds, and ask themselves how much of an investment in analytics tools they are really willing to make.
"There comes a point where you ask how much you're spending to recuperate losses. The question is: how much are you willing to lose?" he said.
This paper seeks to provide education and technical insight to beacons, in addition to providing insight to Apple's iBeacon specification
Focus on cost efficiency, simplicity, performance, scalability and future-readiness when architecting your data protection strategy