ROXANNE

Organised crime is extremely difficult to investigate as law enforcement agencies (LEAs) have to collect, analyse, and connect information from a variety of different sources including eyewitnesses, electronic audio and/or visual monitoring (e.g., wiretaps) and social media.

This environment of super-fast communications, multi-media evidence, and massive volumes of data means the challenge of acquiring and analysing digital evidence is increasingly difficult and extremely time-consuming.

How can we improve the efficiency of organised crime investigations and reduce the burden of organised crime on society?

The ROXANNE project will develop a platform that combines speech, language and video technologies with network analysis tools to assist law enforcement agencies in understanding criminal networks.

The ROXANNE platform includes five interacting components:

• Speaker identification to establish relations between different audio sources, and to potentially determine whether a speaker is among a set of known individuals
• Multilingual speech recognition for rapid and accurate speech-to-text processing of raw audio materials
• Natural language processing to identify entities such as locations, persons and companies from multilingual textual input
• Video and geographical information processing to make use of other visual and spatial information (such as identifying faces, places, backgrounds or geographical position)
• Network analysis to establish connections of these result and enrich them with data from other available sources

Working together, these components provide complementary strengths for investigation and suspect identification. They allow monitoring and reasoning on crimes’ dynamic networks in an automatic manner, so as to speed up the investigative process and reduce the cost and burden to society caused by organised crime activities.

Trilateral assesses the ethical and fundamental rights issues that need to be considered in the development of natural language and video data processing tools for the ROXANNE platform.

Trilateral considers potential risks of data collection, focusing on data collation from criminal proceedings (e.g., wiretap records, associated metadata, other lawfully intercepted data from systems installed in law enforcement agencies), and the measures needed to respect privacy, and ethical considerations to avoid unwanted negative impact to societal values or risks to fundamental rights.

Trilateral engages with law enforcement agencies, policymakers, and civil society organisations to discuss the project framework, gathering feedback on how considerations of societal values, fundamental rights, and applicable legislation can be effectively integrated into the project’s solutions.

Trilateral works on enhancing the project findings by creating a network including law enforcement agencies, researchers, industry, academics, media, policymakers, and civil society organisations to reach out to all relevant stakeholders’ communities.

Our work facilitates novel collaborations, amplifying ROXANNE’s results and encouraging the implementation of the ROXANNE platform by law enforcement agencies.

For more information and updates visit the project website and follow us on Twitter and LinkedIn.