Log in

Combatting new types of fraud: how telecoms operators can use the latest technology

28 June 2017 | Research

Justin van der Lande

Video and podcast | AI and Analytics

New types of fraud are developing in telecoms and known types are becoming more complex. Telecoms operators must take advantage of the latest technology and tools to combat these trends.

Video interview

Justin van der Lande, Principal Analyst at Analysys Mason and Lead Analysts for the Analytics research programme, examines the latest developments in fraud management software for the telecoms industry. 


Listen to and download the podcast


Video transcript 

So one of the areas we have looked at this year is fraud management and fraud management systems on what's going on there. 

There are lots of companies that have been around for some time now addressing fraud in telecoms. 

A traditional way of dealing with this is that you unearth a fraudulant activity, you generate a rule which then looks at various data points to try and find those and you put in processes to try and find those and monitor that activity. That's been pretty successful. 

Generally speaking there have been problems or two challenges which operators come up against . 

Firstly, understanding those frauds tends to happen retrospectively, after the event. They aren't done in real time.  

The second problem is as telecoms gets into new service types, there are new types of services and frauds that are being performed on them.  

So can we use analytics and big data analytics to combat these new types of fraud? 

Well to the first point, absolutely yes. We are seeing a move from batch mode, a retrospective movement in terms of monitoring, into more real-time activity. So you can actually pick up fraudulent activities if you know what it is in terms of the rules, very much quicker. 

The second point is that actually you can use big data analytics to find new types of fraud which a human maybe wouldn't be able to pick up or the complexities of those frauds haven't been able to pick up. 

So you are using a combination of known activities for which you have rules for that you can find quicker and new fraud which can be uncovered by unusual activities and then pinpointed and new algorithms or rules created to go and monitor goes those activities.

So big data is now being used to combat both these thungs within the fraudulant  market space.