Identifying data sources and developing AI-based solutions to analyse online fraud

Addressing online fraud demands a comprehensive understanding of its mechanisms, that relevant data is collected and analysed and that strategies to mitigate exploited vulnerabilities are deployed, alongside public education to reduce the risks of victimisation.

Artificial Intelligence (AI), with its advancements in machine learning, including deep learning and natural language processing, has become a key tool in developing security measures against cybercrime. However, the application of AI in analysing unstructured web data, such as social media conversations and forums for fraud detection, remains underexplored. While previous studies have largely focused on transactional fraud detection for banks, there is a notable gap in understanding and preventing new forms of online fraud using web and police data.

This project brings together researchers from multiple disciplines (computer science/AI, crime science and policing and psychology) from University College London and Anglia Ruskin University to bridge this gap. The project will develop and evaluate AI-based solutions for predicting, detecting, and preventing online fraud, with an emphasis on ensuring these solutions are transparent, reproducible, and interpretable.

Research team

UCL Security and Crime Science (UCL-SCS)

  • Dr Nilufer Tuptuk (Principal Investigator)
  • Dr Enrico Mariconti (Co-Investigator)
  • Dr Antonis Papasavva (Postdoctoral Researcher Fellow)
  • Prof. Shane Johnson (Co-Investigator/Advisor)

UCL Advanced Research Computing (UCL-ARC)

  • Dr Ed Lowton (Co-Investigator)
  • Dr Sanaz Jabbari (Co-Investigator)

Anglia Ruskin University (ARU)

  • Prof Samantha Lundrigan (Co-Investigator)
  • Dr Anna Markovska (Co-Investigator)