Registers – Coded record sensors – Particular sensor structure
Reexamination Certificate
2003-02-25
2004-08-03
Le, Thien M. (Department: 2876)
Registers
Coded record sensors
Particular sensor structure
C235S375000, C235S379000, C235S493000, C705S045000
Reexamination Certificate
active
06769615
ABSTRACT:
TECHNICAL FIELD
The present invention generally relates to a financial document processing system, and, more particularly, to a check processing system.
BACKGROUND OF THE INVENTION
There is a well-defined and well-known process within the banking system of the United States that supports checks as a payment mechanism, commonly known as the check processing or check clearing system.
In a clearing process, a payee deposits a check with the payee's bank. The bank processes the check and forwards it to a clearing agent. The clearing agent then pays the check amount to the payee bank. The types of clearing agents are Federal Reserve Banks, correspondent banks and local clearinghouses (an arrangement whereby a group of banks located in the same geographic area agree to exchange each others checks for presentment at specified times during each business day). The clearing agent physically presents the check to the payer bank and subtracts the check amount from the payer bank. Before the check amount is deducted from the payer's account, the amount, account number, and other important information must be extracted from the check.
The highly automated form of this extraction process performed by banks is done by a check processing control system such as CPCS (Check Processing Control System) made by IBM Corporation of Armonk, N.Y. During the check capture process by the payer bank, the CPCS performs what is known as a prime pass capture of information from the Magnetic Ink Character Recognition (“MICR”) line. The first time a check is passed through a bank's capture equipment, it is referred to as the prime pass capture. The MICR line consists of specially designed numerals (E-13B Font) that are printed on the bottom of a check using magnetic ink. The MICR data fields include the bank routing number, bank transit number, account number, check serial number, check amount, process code and extended process code. Background information related to magnetic ink character recognition systems, and check processing and clearing systems are disclosed in U.S. Pat. No. 5,870,725 issued Feb. 9, 1999 and U.S. Pat. No. 6,243,504 issued Jun. 5, 2001, both of which are incorporated herein in their entirety.
As checks pass through a high speed reader/sorter using CPCS, the magnetic ink is recognized and converted to information that can be used by computers and software for sorting the checks. The reader/sorter is a machine that magnetically recognizes the MICR line on checks to capture the information encoded there, and also uses the information to sort the checks into pre-specified temporary storage slots. This sorting results in a meaningful grouping of the checks that can be used for further processing. Reader/sorters in use at banks today are highly automated capture devices that can capture up to 100,000 checks per hour.
Subsequently, in a process known as balancing, the debits against the accounts associated with the checks are compared with the credits (the payment total made by the bank) to ensure that they are equal.
During the capture process by the high speed reader/sorter, the MICR data is stored in a string known as I-Strings. Each I-String data time or record contains a unique sequence number for each check for each scan. The I-Strings may contain erroneous data from checks that have been rejected by the reader/sorter as being at least partially unreadable. For example, some checks may have partially incorrect routing number due to folding of the check and some checks may have no correct data because they were scanned upside down. The erroneous data for each unreadable check are stored in the I-string and later to a reject string known as RD-String which is created when the CPCS task DIST (Distribution) is run against the I-string. The RD-String contains all of the prime pass reject items.
The erroneous data for the rejected checks can be corrected manually in a process known as Online Reject Reentry (OLRR). In the OLRR process, an operator looks at a screen displaying the rejected item and manually compares it to the actual rejected checks. When the match is found, the operator manually enters the corrected information into a computer. In this case, a simple sequence merge of the manually entered information with the originally scanned data in an I-String is possible because it involves only a single sequence number for each rejected check.
A more efficient method of merging is a high speed reject repair (HSRR) process which allows the bank to recondition the rejected checks and pass them through a high speed reader/sorter a second time to recapture the data in a new string called HSRR I-String. An RR-string is created by merging the HSRR I-string with the repaired rejects from the HSRR capture. Because the rejected checks have been scanned for the second time, the HSRR I-string record for each rejected check receives a new sequence number that is different from the sequence number of the corresponding I-String. Thus, a simple sequence number merge of the RR String data into the corresponding I-String data cannot be done. Moreover, the data in the repaired RR String may look quite different from the corresponding original I-String item.
Consequently, merging of the RR String data with the I-String data in the case of the more efficient HSRR process is a complex procedure that involves matching the individual MICR data fields stored in the two strings. To assist in the merge process, CPCS provides a match profile that is customizable by the user. The match profile typically specifies which MICR fields are to be used during the merge process, how many significant digits need to match, how many digit errors are allowed and how much weight each MICR field should have.
However, the match profile should be set up with great care. If the profile is too strict, much of the data cannot be matched and must be matched by hand in a subsequent balancing process which is a very labor intensive process. On the other hand, if the match profile is too loose, many items that correspond to different checks will be matched incorrectly which results in even more manual work during the subsequent balancing process. As can be appreciated, the profile settings are very sensitive to even slight changes in the matching criteria and are subject to error.
Therefore, there is a need to provide a system and method for an improved merging operation that produces a higher matching rate with less error. It would be desirable to provide such a system with as little change to the bank's current check processing work flow as possible to minimize a user's learning curve.
SUMMARY OF THE INVENTION
According to the principles of the present invention, a novel method and system of processing rejected checks is provided.
Conventionally, a check scanner reads MICR data on the checks which include bank routing information, checking account number, check serial number, check amount, among others. A first pass data includes MICR data of checks that have been scanned for the first time. The second pass data includes MICR data for the rejected checks that have been scanned for the second time, usually after the rejected checks have been reconditioned. In a process known as merging, the second pass data are then compared against the first pass data to find a match according to a match profile which contains the match criteria. The matched data are then passed to a balancing application for final proofing prior to extracting the data for debiting and crediting the accounts involved. However, conventional methods only allowed a single match profile in a one pass merge process.
According to the present invention, a multi-pass merging process is used where the step of comparing the second pass data against the first pass data is performed repetitively using different match profiles. Preferably, a match profile for a subsequent comparison step is less strict than the match profiles of the previous comparison steps. In that embodiment, all previous matches are marked so that they will not be matched with any unlike
Labaze Edwyn
Le Thien M.
Reed Smith LLP
Software Corporation International
LandOfFree
Multi-pass merge process for the check processing control... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Multi-pass merge process for the check processing control..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Multi-pass merge process for the check processing control... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3293893