InfoLab21

Computer experts are now harnessing new developments in language analysis to identify paedophiles posing as children in online chat rooms, to pick up on their vocabulary choices and trail them as they move around the internet.

Computing experts at Lancaster, Swansea and Middlesex Universities have joined forces with specialist UK law enforcement to develop tools to identify paedophiles masquerading as children in online chat rooms.

The social networking boom in recent years means there are more chat rooms and file sharing systems available online than ever before. Paedophiles exploit these environments to groom young users of these sites, to network with other paedophiles and share images of child abuse.

The large volume of child abuse material on the internet presents challenges to those police agencies tasked with monitoring the hundreds and thousands of pages of chat room activity and file downloads that take place. For example, a recent study at Lancaster University found that on the Gnutella peer-to-peer network thousands of searches per second occurred relating to illegal sexual content- including child abuse media.

The researchers will also discover whether new language identification techniques can be used to crack the shared vocabulary of code words used by groups of paedophiles who share files online.

Through testing these monitoring techniques in real-world policing contexts, the technology will be developed as a practical tool for law enforcement officers.

Once developed, these techniques will be automated, potentially freeing up police time and adding to the expertise already deployed within investigations which identify and locate child sex offenders.

This month the University launched Project Isis - a three-year Child Protection Initiative which aims to develop new tools for policing websites and supporting law enforcement. It is funded by the Engineering and Physical Sciences Research Council and The Economic and Social Research Council.

Lead researcher Professor Awais Rashid of Lancaster University's Department of Computing said: "Paedophiles often pose as children online and can be difficult to track as they move quickly from site to site assuming multiple identities.

"Using new language analysis tools and drawing on the expertise of specialist officers, we hope to develop an automated system which can pick up on quirks of language particular to a certain age group. These language patterns can help us to expose adults that seek to groom children online, by posing as children in chat rooms for example.

"The internet also provides other opportunities for paedophiles, such as platforms where they can come together to form secret file-sharing networks. Research shows that people using these networks have a shared vocabulary or code which is constantly evolving. This can make it difficult to gain access to their networks in order to identify the key distributers of child abuse material.

"We hope that using language analysis techniques we can identify these networks and also follow them around the internet even if their shared vocabulary of code words evolves."

He added: "There are privacy and human rights issues involved in any monitoring of this sort. Our goal, therefore, is not only to develop the monitoring techniques to help police apprehend paedophiles but also to study the ethical implications of such monitoring technologies. One-third of the project's activities will focus on ethical considerations and how to incorporate these into the design of the child protection framework."

Thu 19 June 2008

Associated Links

• Computing Department