Method and system for discovering significant subsets in collection of documents

• US 8,118,216 B2
• Filed: 10/30/2007
• Issued: 02/21/2012
• Est. Priority Date: 05/11/2005
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation, JP Morgan Chase Bank, N.A. (JP Morgan Chase & Co)
• Original Assignee: International Business Machines Corporation, JP Morgan Chase Bank, N.A. (JP Morgan Chase & Co)
• Inventors: Hoch, Robert, Ibikunle, Tayo, Kamin, Ehud, Liberis, William A., Nowicki, Tomasz J., Sachar, Howard E., Tresser, Charles P., Walach, Eugene, Reilly, Michael J.
• Primary Examiner(s): HESS, DANIEL A
• Application Number:
  US11/927,977
• Publication Number:
  US 20080061136A1
• Time in Patent Office: 1,575 Days
• Field of Search: 235/379, 235/380, 382/137, 382/139
• US Class Current: 235/379
• CPC Class Codes: G06Q 20/042 characterized in that the payment protocol involves at least one cheque
• CPC Class Codes: G07F 19/00 Complete banking systems; Coded card-freed arrangements adapted for dispensing or receiving monies or the like and posting such transactions to existing accounts, e.g. automatic teller machines data processing equipment for bank accounting G06Q40/02

Abstract

A method (and system) of discovering a significant subset in a collection of documents, includes identifying a set of documents from a plurality of documents based on a likelihood that documents in the set of documents carries an instance of information that is characteristic to the documents in the set of documents.

Claims

1. A method of discovering a subset in a collection of documents, the method being performed on a computer programmed to perform the method, the method comprising:

obtaining a collection of documents, the documents being arranged in sequence and the subset arranged in sequence within the collection of documents;

analyzing a first document in the collection of documents to determine characteristic features of the first document, a plurality of the characteristic features including human created indicia;

generating a profile based on the characteristic features of the first document;

assigning a variable weight to each of the characteristic features;

comparing subsequent documents in the collection of documents to the profile to identify the subset, said comparing comprising considering the characteristic features based on the variable weight assigned to the characteristic features;

during said comparing, redistributing the variable weight when it is determined that one or more of the characteristic features is more reliable than other characteristic fields; and

preselecting a subset of users, said users having created at least one document in the collection of documents, said analyzing is conducted based on said preselecting a subset of users.

2. The method according to claim 1, further comprising:

isolating, after said identifying, said subset of documents from the collection of documents.

3. The method according to claim 2, further comprising:

determining, after said isolating, an identity of said information that is characteristic to the documents.

4. The method according to claim 1,

wherein there is a likelihood that documents included in said subset carry at least one instance of information that is characteristic to said documents in said subset of documents.

5. The method according to claim 1, wherein said cluster of documents comprises a string of documents, said method further comprising determining a beginning and an end of said string of documents, and

wherein said cluster of documents comprises documents other than those randomly spread through said plurality of documents.

6. The method according to claim 1, further comprising:

parameterizing a threshold criteria that determines a significance of said subset of documents.

7. The method according to claim 1, wherein said characteristic feature comprises features selected from a variety of fields on a document.

8. The method according to claim 1, wherein the collection of documents comprises a collection of checks.

9. The method according to claim 1, wherein the human created indicia comprises handwritten features on the document.

10. The method according to claim 1, wherein the characteristic features are unknown prior to said analyzing.

11. The method according to claim 1, wherein the collection of documents comprises a collection of checks, and

wherein the characteristic features identify a payee of the subset of documents in the collection of documents.

12. The method according to claim 1, wherein when said subsequent document matches said profile, said subsequent document is included in said set of documents and a next subsequent document is compared at least to said subsequent document.

13. The method according to claim 1, wherein when said subsequent document does not match said profile, said subsequent document is excluded from said set of documents and a new profile is generated for said subsequent document.

14. A system for discovering a significant subset in a collection of documents, comprising:

an analyzing unit that determines characteristic features from a document in the collection of documents;

a profile-generating unit that generates a profile for said document based on said characteristic features; and

a comparison unit that compares a subsequent document with said profile,

wherein when said subsequent document matches said profile, said subsequent document is included in said set of documents and a next subsequent document is compared at least to said subsequent document.

15. A method, comprising:

analyzing a first document in a collection of documents to determine a characteristic feature of the first document;

generating a profile of the first document based on the characteristic feature; and

comparing a subsequent document in the collection of documents to the profile,

wherein when the subsequent document matches the profile, the subsequent document is included in a set of documents and a next subsequent document is compared at least to the subsequent document.

16. The method according to claim 15, wherein when said subsequent document does not match said profile, said subsequent document is excluded from said set of documents and a new profile is generated for said subsequent document.