Using systems created by Human Inference, the database of ticket applicants was subjected to de-duplication, culturally-aware software and fuzzy logic to root out known hooligans and those who wanted more than the permitted ticket allocation.
Phil Harland, general manager of Human Inference, explained that duplicate names and addresses were filtered out by the de-duplication software. This was also used to remove hooligans by inputting details supplied by the police and various football associations.
Those using changed names or addresses were spotted by software that is aware of cultural differences. This allows for abbreviations of name and address information, or different spellings of similar sounding names.
Fuzzy logic was used to detect subtler changes, so that if someone with a similar name was applying from the same part of a city, the software would come up with less positive responses.
A list of 3,600 potential duplicates was produced from the database of 400,000 applicants, and Euro 2000 refused 1,100 requests.
more e-security news