Use cases

Combating document fraud

Our document fraud detection solution uses advanced data analysis and artificial intelligence to detect anomalies and forgeries, guaranteeing the security and integrity of administrative processes and transactions. The use of this technology is a cornerstone in the fight against document fraud.

Request a demo
Data-Fraud-Detection
Fraud-Detection-AI

The need

Combating document fraud: a priority for government departments!
Our customer, a government ministry, wanted to improve the efficiency of controls on the issuing of administrative documents.
The size of the database (close to a hundred million lines) and the diversity of applications used to enter information - most of which were manual - which had evolved over the last few decades, severely limited the effectiveness of irregularity detection.

Proposed solution

An innovative solution to combat document fraud:

  • Reconciliation of beneficiaries' postal addresses with the BAN - Base Adresse Nationale (National Address Base), resulting in reliable, standardized addresses.
  • Multi-criteria deduplication (surname + first name + address) and multi-strategies (phonetic, Levenshtein distance, N-gram*, etc.) to identify people who have obtained several administrative documents that are supposed to be unique.
  • Blurred joins with other Ministry databases on surname and first name (phonetic + N-gram) as well as on date of birth (with a tolerance of a few days) in order to identify people with several non-cumulative administrative documents.
Data deduplication
Fraud-AI-Tale-of-Data

Gains achieved

Thanks to the use of our document fraud functionalities, the Ministry was able to identify individuals (sometimes up to several hundred in a single department) who possessed several versions of the same administrative document, enabling them to circumvent certain sanctions. The phenomenon had previously gone unnoticed thanks to a few approximations in the spelling of the name, the address (e.g.: street number mentioning the neighboring building) or the date of birth (1 to 2 days apart). To make matters worse, it was impossible to obtain such documents without internal complicity within the Ministry.

Using standardized postal addresses, we were able to identify suspicious addresses used to request a number of administrative documents that far exceeded the number of inhabitants at the address in question (by a factor of 10 or even 100).

By cross-referencing several databases, we were able to uncover numerous cases of unauthorized duplication of administrative documents.

Product Benefits

Ready to elevate your data quality?

Join leading companies improving their data reliability, compliance, and efficiency.
Request a demo