University of Limerick researchers team up with Brazilian police to tackle child porn on dark web

Fintan Walsh

Reporter:

Fintan Walsh

Email:

fintan.walsh@limerickleader.ie

University of Limerick researchers team up with Brazilian police to tackle child porn on dark web

UNIVERSITY of Limerick mathematicians have teamed up with Brazilian police to combat the distribution of child pornography on the dark web. 

Praised by Brazil's Minister for Justice, Sergio Moro, the "innovative collaboration" between the two bodies has sought to make police operations in this area more efficient. 

The research looks into network analysis of Brazil's Operation Darknet, a federal police crackdown on one of the world's largest paedophile online forums. 

These forums are hidden by the Tor browser, an open-source software that allows users to browse the internet anonymously without their IP address being tracked. 

According to the research, published in Nature's Scientific Reports journal, shows that six children were rescued, while 182 users were identified and arrested following the operation between 2014 and 2016. 

One-hundred and seventy of these users were distributors. Among the 10,000 users analysed, 766 were sharing content. 

The team of researchers from the Mathematics Applications Consortium for Science and Industry (MACSI) and the Centre for Social Issues Research, both based at UL, subsequently investigated how effective the operation was in disrupting the distribution of the content.

The research team, which included a Federal Police Officer from Brazil, used the analysis to suggest the best ways to target individuals to maximise the effect of this disruption.

“Network analysis has previously been applied to drug trafficking networks and terrorist networks to identify structural weaknesses and key figures in these illicit networks,” explained Dr Pádraig MacCarron, a postdoctoral researcher at the Centre for Social Issues Research and MACSI who worked with Dr Bruno da Cunha, Kleber A. Oliveira and Professor James Gleeson on the research.

“The dark web network in this study however, was much more dense – as in there were more connections between users than normal - making it more difficult to breakdown using traditional network methods. It was found that the 60% of those core 766 distributors would need to be removed to completely fragment the network. This makes the network highly robust.

“The approach was then taken to try to disrupt the most content shared. Ten of those distributors each had their content viewed more than 100,000 times. This contributed to almost one third of the total views. Of those ten distributors, eight of them were arrested. Of the twenty arrested who had received the most views, they provided 39% of the content. The best the police could have achieved with twenty arrests would have been a reduction of 43%.

“Initially the police investigation was highly effective, quickly arresting those responsible for more than half the content provided. However, subsequent targets were less optimally selected, so that the total arrests reduced the posts by 58% of a possible 92%. This analysis hopes to help lead to more efficient police interventions,” he added.

The team believe the collaboration to be a first between Brazilian law enforcement and Irish mathematicians. Brazilian ministry of justice, Sergio Moro, mentioned the UL research in a recent interview, outlining that “one should invest in intelligence methods to enrich police work” and that “the publication is an index of the quality of the scientific know-how of the Brazilian Federal Police”.

The research was partly funded by Science Foundation Ireland and by the European Research Council.