Method and system for using OCR data for grouping and classifying documents

• US 8,724,907 B1
• Filed: 03/28/2012
• Issued: 05/13/2014
• Est. Priority Date: 03/28/2012
• Status: Active Grant

• Current Assignee: Open Text Corporation
• Original Assignee: EMC Corporation (Dell Technologies Inc.)
• Inventors: Sampson, Steven, Prudent, Yann
• Primary Examiner(s): Li, Ruiping
• Application Number:
  US13/432,251
• Time in Patent Office: 776 Days
• Field of Search: 382/218, 382/209, 382/224, 382/177, 707/999.006, 707/E17.008, 707/737
• US Class Current: 382/218
• CPC Class Codes: G06F 16/93 Document management systems
• CPC Class Codes: G06F 18/23 Clustering techniques
• CPC Class Codes: G06F 18/2433 Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
• CPC Class Codes: G06F 18/28 Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/153 using recognition of characters or words
• CPC Class Codes: G06V 30/1914 Determining representative reference patterns, e.g. averaging or distorting patterns; Generating dictionaries, e.g. user dictionaries
• CPC Class Codes: G06V 30/19173 Classification techniques
• CPC Class Codes: G06V 30/412 Layout analysis of documents structured with printed lines or input boxes, e.g. business forms or tables
• CPC Class Codes: G06V 30/418 Document matching, e.g. of document images

Abstract

A document template for classifying documents is created for each document class. The document template includes a set of keywords and the spatial relations of the keywords. A document to be classified is received. The spatial relations of the template keywords of a template are compared with the spatial relations of corresponding words in the document. If the spatial relations are the same, the document may be classified in the document class of the template.

Claims

1. A system for classifying digitized documents, the system comprising:

a processor-based document management system executed on a computer system and configured to;

create and store a plurality of templates associated with a plurality of document classes, each template comprising a plurality of keywords;

receive a digitized document to be classified;

compare each template with the digitized document to be classified, wherein the comparison comprises;

comparing a first area value associated with a template with a second area value associated with the digitized document,

the first area value associated with a keyword indicating an area occupied by the keyword in the template, and

the second area value that indicates an area occupied by a word in the digitized document to be classified;

determine that a difference between the first and second area values is below a threshold value; and

upon the determination that a difference is below a threshold value, identify the keyword as being a keyword for a word pair, and identify the word in the digitized document to be classified as being a corresponding word for the word pair.

2. The system of claim 1 wherein the processor-based document management system is configured to match of words of the digitized document with the keywords of the template.

3. The system of claim 1 wherein the second location information comprises top location information, and bottom location information,

the top location information is associated with a top portion of the digitized document, and comprises a location of a word in the top portion of the digitized document relative to other words in the top portion, and

the bottom location information is associated with a bottom portion of the digitized document and comprises a location of a word in the bottom portion of the digitized document relative to other words in the bottom portion.

4. The system of claim 1 wherein the processor-based document management system is configured to:

calculate a first vector from the keyword in the template to another keyword in the template, the first vector thereby indicating a location of the keyword relative to the other keyword; and

calculate a second vector from the word in the digitized document to another word in the digitized document, the second vector thereby indicating a location of the word in the digitized document relative to the other word in the digitized document.

5. The system of claim 1 wherein the processor-based document management system is configured to:

calculate a horizontal distance from the keyword in the template to another keyword in the template; and

calculate a vertical distance from the keyword in the template to the other keyword in the template, the horizontal and vertical distances thereby indicating a location of the keyword relative to the other keyword.

6. The system of claim 1 wherein the plurality of templates are stored before the digitized document to be classified is received.

7. The system of claim 1 wherein the processor-based document management system is configured to:

calculate a Levenshtein distance between a keyword from the template and a word from the digitized document;

determine that the Levenshtein distance is below a threshold value; and

upon the determination, identify the keyword from the template as being a word for a word pair, and identify the word from the digitized document as being a corresponding word for the word pair.

8. A method comprising:

creating and storing a plurality of templates associated with a plurality of document classes, each template comprising a plurality of keywords;

receiving a digitized document to be classified;

comparing each template with the digitized document to be classified, wherein the comparison comprises;

comparing a first area value associated with a template with a second area value associated with the digitized document,

the first area value associated with a keyword indicating an area occupied by the keyword in the template, and

the second area value that indicates an area occupied by a word in the digitized document to be classified;

determine that a difference between the first and second area values is below a threshold value; and

upon the determination that a difference is below a threshold value, identify the keyword as being a keyword for a word pair, and identify the word in the digitized document to be classified as being a corresponding word for the word pair.

9. The method of claim 8 comprising matching keywords of the template with words of the digitized document.

10. The method of claim 8 wherein the second location information comprises top location information, and bottom location information,

the top location information is associated with a top portion of the digitized document, and comprises a location of a word in the top portion of the digitized document relative to other words in the top portion, and

the bottom location information is associated with a bottom portion of the digitized document and comprises a location of a word in the bottom portion of the digitized document relative to other words in the bottom portion.

11. The method of claim 8 comprising:

calculating a first vector from the keyword in the template to another keyword in the template, the first vector thereby indicating a location of the keyword relative to the other keyword; and

calculating a second vector from the word in the digitized document to another word in the digitized document, the second vector thereby indicating a location of the word relative to the other word.

12. The method of claim 8 comprising:

calculating a horizontal distance from the word in the digitized document to another word in the digitized document; and

calculating a vertical distance from the word in the digitized document to the other word in the digitized document, the vertical and horizontal distances thereby indicating a location of the word relative to the other word in the digitized document.

13. The method of claim 8 wherein the plurality of templates are stored before the digitized document to be classified is received.

14. The method of claim 8 comprising:

calculating a Levenshtein distance between a keyword from the template and a word from the digitized document;

determining that the Levenshtein distance is below a threshold value; and

upon the determination, identifying the keyword from the template as being a word for a word pair, and identifying the word from the digitized document as being a corresponding word for the word pair.

15. A computer program product, comprising a non-transitory computer-readable medium having a computer-readable program code embodied therein, the computer-readable program code adapted to be executed by one or more processors, the program code including instructions to:

create and store a plurality of templates associated with a plurality of document classes, each template comprising a plurality of keywords;

receive a digitized document to be classified;

compare each template with the digitized document to be classified, wherein the comparison comprises;

comparing a first area value associated with a template with a second area value associated with the digitized document,

the first area value associated with a keyword indicating an area occupied by the keyword in the template, and

the second area value that indicates an area occupied by a word in the digitized document to be classified;

determine that a difference between the first and second area values is below a threshold value; and

upon the determination that a difference is below a threshold value, identify the keyword as being a keyword for a word pair, and identify the word in the digitized document to be classified as being a corresponding word for the word pair.

16. The computer program product of claim 15 wherein the second location information comprises top location information, and bottom location information,

the top location information is associated with a top portion of the digitized document, and comprises a location of a word in the top portion of the digitized document relative to other words in the top portion, and

the bottom location information is associated with a bottom portion of the digitized document and comprises a location of a word in the bottom portion of the digitized document relative to other words in the bottom portion.

17. The computer program product of claim 15 wherein the method comprises:

calculating a first vector from a keyword in the template to another keyword in the template, the first vector thereby indicating a location of the keyword relative to the other keyword; and

calculating a second vector from the word in the digitized document to another word in the digitized document, the second vector thereby indicating a location of the word relative to the other word.

18. The computer program product of claim 15 wherein the method comprises:

calculating a horizontal distance from the word in the digitized document to another word in the digitized document;

calculating a vertical distance from the word in the digitized document to the other word in the digitized document, the vertical and horizontal distances thereby indicating a location of the word relative to the other word in the digitized document.