Littoralis Limited is working with members of the National Association of Business Crime Partnerships (NABCP) to create a national information-sharing network.
The system, based on Littoralis’ DISC online partnership management system, will help identify travelling, prolific and professional shoplifters and teams.
DISC is in use in more than 140 crime reduction partnerships, covering 260 or more towns and city-centres. The new ‘Data Sharing Group’ will enable the administrators of those partnerships which are NABCP members to interrogate each other’s data to find close matches to any of their own local offenders.
All participating NABCP partnerships are accredited with the national standard, the Safer Business Accreditation, and work within the Data Protection Act; the NABCP Data Sharing Group itself will operate in strict compliance with the DPA, organisers stress.
Administrators can only seek matches against individual offenders in their own databases; ‘fishing’ for personal data is not possible in the Data Sharing Group. Where a possible match has been identified, each administrator must agree to their own satisfaction that the two data items refer to the same individual before agreeing between themselves to share the specific items.
Data Sharing Groups, police force-wide, began in Hampshire in April and since others have been deployed across Thames Valley, Sussex, Devon and Cornwall, Avon and Somerset, Cheshire, West Yorkshire and Gloucestershire. The largest, until the launch of the NABCP’s, was the London-wide group linking some 20 partnerships.
Charlie Newman of Littoralis, pictured, says: “The one-click matching process applies a complex search algorithm across the combined databases, and uses a large number of search criteria. Obviously the offender’s name is important – if it’s available – but the matching compares build, gender, ethnicity, any other visual characteristics, the type of offences associated with the individuals and any other information that may be available including address, data of birth and so on. Not all this information is always available of course, but the algorithm simply uses everything known about an offender to fetch a list of possible matches, ranked by similarity.
“The beauty of the system is that it doesn’t rely on still-unproven automated facial recognition systems which only work tolerably well with clearly-lit, well-defined, full-face images – and most CCTV images are very far from that. The Cross-DISC Data Sharing system simply selects the top-15 best-matches and enables the administrator to visually compare them to see if any of them are, indeed, the same individual.”
Littoralis hope to sponsor a conference on the subject, provisionally titled ‘Dare to Share’, in the first quarter of 2017, aimed at UK police forces.
