2.01 Detecting grooming behaviour on social media
Academic team: Prof Harith Alani, Dr Elizabeth Cano, Dr Miriam Fernandez,
Policing partners: Dorset Police, Avon and Somerset Police, Lancashire Constabulary
Status: Complete
Online paedophile activity has become a major concern in society with the internet widely available to the general population and young people.
This piece of research looked into whether the different stages of online grooming behaviour could be automatically detected. The proposed approach combines Machine Learning (ML) techniques with existing psychological theories and discourse studies to better encapsulate existing knowledge of online grooming.
The results of this study demonstrate the effectiveness of this approach for the automatic detection of online grooming stages, opening new possibilities for addressing predator grooming behaviour online, and helping policing organisations to act in a preventive way.
Outputs
| Title | Outputs type | Lead academic | Year |
|---|---|---|---|
| Detecting child grooming behaviour patterns on social media | Executive summary | Alani, H | 2017 |
| Detecting grooming behaviour on social media | Presentation | Alani, H | 2017 |
| Detecting child grooming behaviour patterns on social media | Final report | Alani, H | 2016 |
News
Findings from an evaluation of the pilot application of AI for witness statement and report generation
Dr Paul Walley and Dr Helen Glasspoole-Bird have published an evaluation report entitled “An Evaluation of the Pilot Application of Artificial Intelligence to Witness Statement and Report Generation at Hertfordshire Constabulary”. The work studies the outputs of version 1 of an AI application that takes audio from Rapid Video Response interviews with victims of domestic abuse and converts this into relevant summary documents including MG11 witness statements.
Explore
Undergraduate
Postgraduate
Policy
-
Follow us on Social media
-
