European Commission logo
English English
CORDIS - EU research results

Retrieval and Analysis of Heterogeneous Online Content for Terrorist Activity Recognition

Article Category

Article available in the following languages:

New tools help law enforcement stop online terrorist activity

The TENSOR project is developing an integrated intelligence platform that will help Law Enforcement Agencies quickly detect online terrorist activities, radicalisation and recruitment.

Security icon Security

For the vast majority of us, the internet is a valuable resource and an ingrained part of our everyday lives. But for criminals and terrorists, the internet is a tool for communicating, coordinating, recruiting and fundraising. As these activities present a significant risk to the safety of European citizens, it is essential that Law Enforcement Agencies have the tools they need to identify, gather and interpret terrorist-generated online content. This is why the EU-funded TENSOR project is developing an integrated terrorism intelligence platform that will help law enforcement quickly detect online terrorist activities, radicalisation and recruitment. “The core objective of the project is to keep people safe from the ever-present threat of terrorism,” says Jonathan Middleton, TENSOR project coordinator. “This means not only providing Law Enforcement Agencies with effective solutions for preventing terrorism, but also ensuring that these solutions comply with the privacy laws that protect an individual’s freedom to use the internet.”

Automated and semi-automated tools

The TENSOR platform integrates a range of automated and semi-automated tools, including: targeted terrorist content discovery, information extraction, content categorisation, filtering and analysis technologies; real-time content summarisation and visualisation; automated audit trails, and privacy-by-design and data protection components. These tools then search, crawl and monitor the entire web – from the Dark Web to social media, multimedia and even multilingual content – looking for possible terrorist-generated content. The platform then extracts and analyses all relevant information, the results of which are provided to law enforcement via an intuitive user interface and in a highly visual format. “The TENSOR System is coupled with engineering and customised services that meet law enforcement needs and existing investigative workflows,” explains Middleton. “In this sense, it is an organic system that emphasises customisation and integration to ensure its compatibility with a wide range of existing systems and tools.” Although the platform is currently in the last stages of development, several important results have already been achieved. For example, by linking multiple pieces of data and users, TENSOR can semi-automatically inform the intelligence officer or analyst of a potential terrorist connection. The platform is also showing early success in its ability to provide real-time analysis and translation of multilingual social media and web content.

Keeping people safe

As TENSOR nears completion, researchers are exploring potential opportunities for exploiting the results and innovations created during the project. Already, a number of Law Enforcement Agencies have expressed interest in working with the project partners to further develop and deploy the platform. “By bringing together Law Enforcement Agencies, academia and private sector businesses, we have developed a tool that meets today’s intelligence needs and, most importantly, keeps people safe,” adds Middleton. “As a result, TENSOR’s legacy will be its contribution to helping law enforcement tackle the impact of terrorist-generated online content.” Whilst the focus of the TENSOR project is the discovery of terrorist-generated online content, researchers note that there is scope to adapt the platform to detect other areas of criminality. This potential will be explored following the project’s completion.


TENSOR, law enforcement agencies, terrorist, radicalisation, internet, terrorism, privacy, online content

Discover other articles in the same domain of application