Method for identifying suspect items in an out-of-balance transaction

• US 6,125,196 A
• Filed: 02/25/1994
• Issued: 09/26/2000
• Est. Priority Date: 10/02/1992
• Status: Expired due to Term

• Current Assignee: Unisys Corporation
• Original Assignee: Unisys Corporation
• Inventors: Carey, Daryl James, Muszynski, Daniel W., Ladouceur, Lynn Leppek
• Primary Examiner(s): MEHTA, BHAVESH M
• Application Number:
  US08/202,860
• Time in Patent Office: 2,405 Days
• Field of Search: 382/1, 382/7, 382/57, 382/137, 382/138, 382/139, 382/140, 235/379, 235/425, 235/437, 235/470, 364/400, 364/404, 364/405, 364/406
• US Class Current: 382/138
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/127 with the intervention of an operator
• CPC Class Codes: G06V 30/274 Syntactic or semantic context, e.g. balancing

Abstract

A method for detecting the most likely source of an error in a document image processing system which processes sets of related documents. A set of characteristics is selected as being indicative that the document image, or item, is the source of the error. Each of the characteristics is assigned a weight, wherein the weight is in proportion to the likelihood that an item exhibiting the characteristic is the source of error. The items are ranked according to the characteristics exhibited by each item.

Claims

1. An improved method for balancing an out-of-balance transaction in a check image processing system, wherein one or more of the items in the out-of-balance transaction has an associated amount and may be the source of the error, the method comprising the steps of:

associating a different one of a plurality of weights with each of a plurality of different characteristics, wherein each of said plurality of different characteristics indicates a different type of error which may be causing the out-of-balance condition, and each of said plurality of weights indicates a different relative likelihood that said each of said plurality of different characteristics from said associating step is the source of the error in the out-of-balance transaction, and wherein said associating step includes the step of classifying said different characteristics into a first set of characteristics and a second set of characteristics, wherein said first set of characteristics is indicative of a greater likelihood of error than said second set of characteristics, and each of said plurality of different characteristics in said first set is descriptive of amounts which may be partially correct and each of said plurality of different characteristics in said second set is descriptive of the physical appearance of the amounts;

identifying which of said plurality of different characteristics are exhibited by each of the items;

detecting which ones of the items are suspect items, wherein a suspect item is an item for which at least one of said plurality of different characteristics from said identifying step was identified;

ranking each of said suspect items to indicate the relative likelihood that each of said suspect items is the source of the error, wherein the relative likelihood of a suspect item being the source of the error is determined by a characteristic exhibited by said suspect item and said one of said plurality of weights associated with said characteristic;

reentering the amounts for the items in the out-of-balance transaction until the transaction balances, wherein said reentering step begins with said suspect item which is the most likely source of the error and progresses toward the item which is the least likely source of error; and

checking whether the transaction is balanced after each amount is reentered, whereby said reentering step is completed upon detection of the transaction being balanced.

2. The method of claim 1, wherein said first set of characteristics includes a significant-digit characteristic, a highest-risk-digit characteristic, an extra-digit characteristic, a common-MICR-misread characteristic, and a transposition characteristic.

3. The method of claim 2, wherein said second set of characteristics includes a hand-written characteristic, a machine-printed characteristic, a complex-cent characteristic, a dash-trailing characteristic, and a touching-characters characteristic.

4. An improved apparatus for balancing an out-of-balance transaction in a check image processing system, wherein one or more of the items in the out-of-balance transaction has an associated amount and may be the source of the error, comprising:

association means for associating a different one of a plurality of weights with each of a plurality of different characteristics, wherein each of said plurality of different characteristics indicates a different type of error which may be causing the out-of-balance condition, and each of said plurality of weights indicates a different relative likelihood that said each of said plurality of different characteristics is the source of the error in the out-of-balance transaction, said association means including classification means for classifying said plurality of different characteristics into a first set of characteristics and a second set of characteristics, wherein said first set of characteristics is indicative of a greater likelihood of error than said second set of characteristics, and each of said plurality of characteristics in said first set is descriptive of amounts which may be partially correct and each of said plurality of characteristics in said second set is descriptive of the physical appearance of the amounts;

identification means for identifying which of said plurality of different characteristics are exhibited by each of the items;

detection means responsive to said identification means for detecting which ones of the items are suspect items, wherein a suspect item is an item for which at least one of said plurality of different characteristics was identified by said identification means;

ranking means responsive to said detection means and interfaced with said plurality of weights for ranking each of said suspect items to indicate the relative likelihood that each of said suspect items is the source of the error, wherein the relative likelihood of a suspect item being the source of the error is determined by a characteristic exhibited by said suspect item and said one of said plurality of weights associated with said characteristic;

presentation means responsive to said ranking means for presenting for reentry the amounts for the items in the out-of-balance transaction, wherein said suspect items are presented beginning with said suspect item which is the most likely source of the error and progressing toward the item which is the least likely source of error; and

checking means responsive to amounts reentered and interfaced with said presentation means for checking whether the transaction is balanced after each amount is reentered and signaling said presentation means that presentation of suspect items is complete.

5. The apparatus of claim 4, wherein said first set of characteristics includes a significant-digit characteristic, a highest-risk-digit characteristic, an extra-digit characteristic, a common-MICR-misread characteristic, and a transposition characteristic.

6. The apparatus of claim 5, wherein said second set of characteristics includes a hand-written characteristic, a machine-printed characteristic, a complex-cent characteristic, a dash-trailing characteristic, and a touching-characters characteristic.