Data mining key to revealing hidden criminal behaviours, says SAS

graph1SAS Public Security says it is targeting an emerging market for security which fits between the traditional private sector and government sectors.

According to the company: “Law enforcement, security and border agencies the world over face a raft of challenges – from dealing with huge amounts of data that needs to be entered into disparate systems and greater demand on them for accurate and timely information through to sharing information in a rapid and efficient way with other agencies and departments on a regional and global scale. SAS has solutions in place that address these challenges, ensuring that information is turned into ‘actionable intelligence’ that enables fast responses and effective action.”

Ian Manocha, Vice President of Government for SAS Europe Middle East and Africa

Ian Manocha, Vice President of Government for SAS Europe Middle East and Africa

In a seminar session at its UK headquarters in Marlow, Ian Manocha, vice president of government for SAS Europe Middle East and Africa estimated the value of this new sector is £157 billion.

Budget cutting, which will force security departments and organisations to cut manpower and make better use of their resources, combined with information overload, has created the conditions which are ripe for SAS’s product offering – namely, software solutions which aim to improve intelligence processes, enhance public safety and prevent and deter crime and terrorism.

Joanne Taylor, Director, SAS Public Security

Joanne Taylor, Director, SAS Public Security

Joanne Taylor, director of public security at SAS, painted a picture of an evolving threat. Twenty years ago, the IRA was bombing civilian targets but the threat was localised and could be understood in terms of what was happening in Northern Ireland.

After 9/11, the threat became much harder to analyse. “We’ve got better at filling in the intelligence gaps but now there are multiple targets, potentially thousands of people who pose a varying level of threat,” she said. “Do we run a full scale surveillance op on them, which costs a lot, or are they just making a lot of noise? We have to make a decision about where to apply resources.

“This doesn’t just apply to terrorism, it can also be applied to other crimes like sex offenders. In one medium sized force, they are tracking 7,000 sex offenders and they only employ 10,000 police officers.”

The sheer amount of information that has to be analysed is making the job more difficult, she said. In 2009, there was something on the order of 0.8 zettabytes of data (a zettabyte is one billion terabytes or one trillion gigabytes) recorded on hard drives around the world. Last year, the figure had risen to 35.2 zettabytes, an increase of 44 times in just three years.

slide2

Taylor suggested that “mining” ANPR data – of which there are billions of records – would be useful in combatting crime. Combined with other sources of data, this could provide a rich source of intelligence for police and the security services.

However, as we have learned in recent years, information overload is a serious issue, and how do you run a search for something when you don’t necessarily know what you are looking for?

“This is where advanced analytics come in. it’s for when you don’t know what the question should be. It pushes information out to the human user who can then say, yes, we need to look at that question that has been raised by the machine,” she said.

Historically, analysts have used link analysis to examine data.

slide3

However, as data becomes more complex, you need to shift to “sequence analysis” to understand, for instance, how someone become radicalised to become a terrorist or perhaps how someone moves toward becoming a sex offender.

Machines which can analyse data and extract information that points to these types of behaviour would be a big help to the police and intelligence services, she said. The challenge is dealing with the heaps of data, 80 per cent of which is unstructured, that is to say, is mostly just text.

To analyse this, you need to have machines that can interpret the meaning of text and understand the use of words in context.

According to SAS, its software can analyse data from numerous sources such as blogs and twitter and alert human operators to suspicious activity.

For more information on SAS Public Security, visit the company’s website.

Links
SAS Public Security 

Leave a Comment

You must be logged in to post a comment.