Theme-based system and method for classifying documents

• US 7,376,635 B1
• Filed: 07/21/2000
• Issued: 05/20/2008
• Est. Priority Date: 07/21/2000
• Status: Expired due to Term

• Current Assignee: Ford Global Technologies, LLC (Ford Motor Company)
• Original Assignee: Ford Global Technologies, LLC (Ford Motor Company)
• Inventors: Montgomery, Thomas Anthony, Goodman, Bryan Roger, Porcari, Damian
• Primary Examiner(s): RIMELL, SAMUEL G
• Application Number:
  US09/621,393
• Time in Patent Office: 2,860 Days
• Field of Search: 707/5, 707/6, 707/104.1, 707/1, 707/7, 707/3, 715/513, 715/500, 715/514
• US Class Current: 1/1
• CPC Class Codes: G06F 16/355 Class or cluster creation or modification
• CPC Class Codes: G06F 40/35 Discourse or dialogue representation
• CPC Class Codes: Y10S 707/99931 Database or file accessing

Abstract

A classification system (10) having a controller (12), a document storage memory (14), and a document input (16) is used to classify documents (20). The controller (12) is programmed to generate a theme score from a plurality of source documents in a plurality of predefined source documents. A theme score is also generated for the unclassified document. The unclassified document theme score and the theme scores for the various classes are compared and the unclassified document is classified into the classification having the nearest theme score.

Claims

1. A method for classifying a document using a classification system comprising the steps of:

defining a plurality of classes;

identifying source documents of each of said plurality of classes;

generating a classification theme score for each of said classes from the source documents for each of the plurality of classes;

entering an unclassified document into the system;

generating an unclassified document theme score corresponding to said unclassified document;

classifying the unclassified document into one of said plurality of classes when the unclassified document theme score is substantially equal to the classification theme score;

reclassifying a plurality of classes into a plurality of new classes by;

identifying source documents for each of said plurality of new classes;

generating a respective plurality of new class theme scores for each of said plurality of new classes;

reclassifying documents within said plurality of classes into the plurality of new classes when a classified document theme score is substantially similar to one of the respective new class theme scores;

storing the reclassified documents in a document storage memory.

2. A method for classifying a document using a classification system comprising the steps of:

establishing a plurality of classes and a plurality of subclasses;

identifying source documents of each of said plurality of classes and said plurality of subclasses;

generating a classification theme score for each class and each subclass in response to the source documents;

entering an unclassified document into the system;

generating an unclassified theme score for the unclassified document; and

classifying the document into one of said plurality of classes when the unclassified document theme score is substantially equal to the classification theme score;

reclassifying a plurality of classes into a plurality of new classes by;

identifying source documents for each of said plurality of new classes;

generating respective new class theme scores;

reclassifying documents with said plurality of classes into the plurality of new classes when the classified document theme score is equal to one of the respective new class theme scores;

storing the reclassified documents in a document storage memory.

3. A system for classifying documents comprising:

a document input for entering an unclassified document;

a document storage memory;

a controller coupled to said document input and document storage memory said controller programmed to classify documents into a plurality of classes by identifying source documents of each of said plurality of classes, generating a classification theme score categorizing documents into said classes, generating an unclassified theme score for the unclassified document, and classifying the document into one of said plurality of classes when the unclassified document theme score is substantially similar to the classification theme score, said controller further programmed to identify source documents for each of a plurality of new classes;

generate respective new classification theme scores;

reclassify documents with said plurality of classes into the plurality of new classes when the classified document theme score is substantially similar to the new class theme score; and

store the reclassified documents in the document storage memory.

4. A system as recited in claim 3 wherein said controller is programmed to assign a weight to sections of an unclassified document.

5. A system as recited in claim 3 wherein said controller is programmed to classify documents into a subclass.

6. A system as recited in claim 3 wherein said document input comprises the Internet.

7. A system as recited in claim 3 wherein said document input comprises a scanner.