A Queensland-based statistics researcher is working to counter violent extremism by examining the way people talk about themselves and their beliefs.

“Examining what they say about themselves and what they believe could help agencies identify at-risk individuals and help prevent them from turning to violence,” says Dr Gentry White, a lecturer in statistical science in QUT's Science and Engineering Faculty.

He is using maths to develop a quantitative model for analysing interview transcripts of recent converts and radicalised followers in Canada, US the UK and Australia, in a project funded by the US Department of Defence.

“The model will help counter terrorism experts look for any similarities between the trajectories to conversion and radicalisation,” Dr White said.

“Participants are asked about how they got interested in Islam, why they converted or about the process of radicalisation. Some themes around family and sense of belonging are emerging.”

Dr White has also analysed terrorism incidents from the Global Terrorism Database to find and examine patterns of terrorist activity and identify key terrorist events.

“If we find clusters, mathematical models are useful in discerning patterns of behaviour,” he said.

“If we look at the parameters that describe the clusters in these countries, we notice they change from country to country so we can qualify how different conflicts behave.

“We are able to look for signature events and patterns over time and they vary from conflict to conflict.

“Most events have a very minimal excitation effect, that is most are one-off sporadic events.

Cyclical patterns have emerged because after a major event there is a crackdown by authorities and terrorist groups go underground into a passive planning cycle.

“My most recent research shows most events are not harbingers of immediate future attacks,” Dr White says.

“There is a subset of events, however, that are influential and we have to analyse their characteristics to find what distinguishes them.

“The main feature consistent across all terrorist contexts is that there are a small number of extremely influential events and a large number of events that don't have much influence.

“In Colombia, for example, we found the event that was influential in starting a string of attacks on politicians to manipulate election outcomes.”