Browse Prior Art Database

Check Fraud Detection System For Financial Institutions

IP.com Disclosure Number: IPCOM000013524D
Original Publication Date: 2001-May-01
Included in the Prior Art Database: 2003-Jun-18
Document File: 2 page(s) / 57K

Publishing Venue

IBM

Abstract

Disclosed is a system that allows for automated check and currency fraud detection. Once integrated with a check/currency sorter, it allows for outsorting of suspect items, based on fraud (logical and physical) detection techniques.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 98% of the total text.

Page 1 of 2

Check Fraud Detection System For Financial Institutions

  Disclosed is a system that allows for automated check and currency fraud detection. Once integrated with a check/currency sorter, it allows for outsorting of suspect items, based on fraud (logical and physical) detection techniques.

The logical system is a knowledge based system and suggest outsorting of checks based on criterion, customized to accounts, criterion include:

1. Signature verification accounts.
2. High risk accounts.
3. High amount.
4. New account.
5. Account history:
a. Large fluctuating balance.
b. Frequent overdrafts.
c. Current balance.
6. Account classification.
a. Personal.
b. Small/large business.
7. Illogical transaction.
a. Payer to payee.
8. Greatly out of sequence checks.
9. Invalid account or branch number.
10. Authorized document pay/no pay.
11. Recent history of paid code lines.
12. Document date.

   Criterion based on physical detection are:

1. Presence of UV ink on checks.
2. Non magnetic E13B characters in the code line.
3. Texture analysis.
4. Image analysis based on high resolution scanning.
5. Courtesy amount alteration.
6. Payee line alteration.
7. Checks physical attributes (size) etc different than nominal.
8. Mismatch of legal and courtesy amount.

Specifically applicable to currency are:

1. Verification of embedded images.
2. Verification that fine printing details are not missing from the scanned image.
3. Verification that coloured portions of the image have not lost their col...