Menu
a chain fence with lights behind it.
Algorithms – digital detectives in the fight against terrorism

The following article is written by algorithms and cybercrime expert Dr Nick Patterson from Deakin University’s School of Information Technology.

With the recent terrorist attacks in the UK and France, it’s natural for Australians to worry whether we might be in danger too. In the past two years, Australia has allegedly avoided 15 terror attacks. However, cyber security has taken counter-terrorism efforts to another level and, excitingly, computer algorithms are helping protect Australia by flagging the crime before it even takes place.

It’s likely that many of the 15 foiled terrorist attacks may have been prevented by algorithmic solutions. As former Australian Crime Commission (ACC) Chief Information Officer Maria Milosavljevic says, ‘the ACCC was using aspects of data mining to trawl through data sets looking for patterns and potentially predicting emerging crime issues and trends across the country‘.

Identifying terrorists using algorithms

One of the keys to stopping terror attacks in Australia is utilising digital solutions, which are used to zero-in on anomalies like unusual stopovers in known terrorist hotspots, biometric data such as fingerprints, and last-minute ticket purchases, in order to identify terrorists trying to enter the country.

To identify potential terrorists, Australia, the United States, the United Kingdom, Canada and New Zealand employ an intelligence network and security arrangement system called Five Eyes. It’s likely (top secret!) the network uses a computer algorithm in coordination with large shared databases to look for red flags – usually in travel data.

There are many other factors that can be analysed too. Here, I’ve narrowed down seven different red flags, which would likely be seen in a predictive algorithm operated by Five Eyes:

 

  1. Departing or visiting countries which are known for active terror attacks
  2. Someone other than yourself purchasing your tickets
  3. Last-minute plane ticket purchases and itinerary changes
  4. Short stays in a foreign country (red flags if this country is known for its terrorist activity)
  5. Unusual travel routes, stopping off at countries for a short time
  6. Known relationships or related to convicted terrorists or terrorist groups
  7. Biometric data like fingerprints, physical features (e.g. length of nose) and iris scans.
'One of the keys to stopping terror attacks in Australia is utilising digital solutions.'

Dr Nick Patterson,
School of Information Technology, Deakin University

How do algorithms work?

Most likely if you’re not into maths or programming you won’t know what an algorithm is. An algorithm is essentially a set of instructions that takes input, conducts some processing on this input whereby the data is often changed, and then a result is determined at the end.

How successful are algorithms?

The 2015 government review of counter-terrorism machinery stated the number of people in Australia who are prepared to commit terror attacks is quite low, which is fantastic. However, this isn’t just by magic; it has been achieved by authorities through travel restrictions. A recent example of the red flag in action saw more than more than 500 Iraqi and Syrian refugees bound for Australia in 2017 refused entry after their names were found on watch lists.

These types of algorithmic-based systems are capable of scanning through monumental amounts of civil, surveillance and criminal big data efficiently and effectively. Algorithms do the work of potentially thousands of police detectives, meaning much more new leads and evidence can be gathered to support criminal investigations, which in turn can prevent crime that would normally go unnoticed.

Algorithms used around the world

There are many other successful algorithm-based solutions such as PredPol, which crunches through data to predict crimes. The Los Angeles Police Department found, through using this, they had a 20 per cent drop in predicted crimes from 2013 to 2014.

In Australia, we are also now able to predict crime trends through the Australian Crime Commission’s Project Fusion, which has been developed over four years utilising a $14.5 million budget.

Project Fusion uses crime-prediction models by integrating big data and using predictive analytics to scan for patterns, then automating certain tasks that a human analyst would normally do. Algorithms are making resource-heavy tasks a thing of the past.

Solutions such as PredPol and Project Fusion are great additions to the arsenal to fight modern crime, and there’s still scope for greater advancements in the area. This will come from governments forcefully gaining access to our private information. Where governments run into trouble is that most of us are usually unwilling to lose our privacy in the name of security.

The benefits of predictive policing and algorithms are vast, but we can’t stop every crime or terrorist act just yet. No authority or system was able to prevent the ISIS terror attack in Melbourne in June, due to a stifled legal system. The suspect had a long criminal history, which should have been an obvious red flag.

It’s clear that algorithms have a huge part to play in understanding the threats and challenges the world faces in counter-terrorism security from the UK and France to Australia and New Zealand.

Interested in a career in data security? Check out Deakin’s range of data analytics and cyber security courses.

this. featured experts
Dr Nick Patterson
Dr Nick Patterson

School of Information Technology, Deakin University

Read profile

explore more