Index extraction from documents

• US 20060036649A1
• Filed: 08/12/2004
• Published: 02/16/2006
• Est. Priority Date: 08/12/2004
• Status: Abandoned Application

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Simske, Steven J., Wright, David W.
• Application Number:
  US10/916,878
• Publication Number:
  US 20060036649A1
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/353 into predefined classes
• CPC Class Codes: G06F 16/93 Document management systems

Abstract

Systems, methods, and programs embodied in a computer readable medium are provided for index extraction. A plurality of ground truth documents are stored in a database, the ground truth documents being organized in a plurality of classifications. Attempts are made to automatically extract indices from a document based upon a classification associated with the document. The document is reclassified from a first one of the classifications to a second one of the classifications during the course of the automated extraction of the indices by drawing an association between the document and at least one of the ground truth documents. The indices are manually extracted from the document upon a failure to automatically extract the indices. The document is stored in the database as one of the ground truth documents if the indices are manually extracted.

Claims

1. A method for indexing documents, comprising the steps of:

storing a plurality of ground truth documents in a database, the ground truth documents being organized in a plurality of classifications;

attempting to automatically extract indices from a document based upon a classification associated with the document;

reclassifying the document from a first one of the classifications to a second one of the classifications during the course of the automated extraction of the indices by drawing an association between the document and at least one of the ground truth documents;

manually extracting the indices from the document upon a failure to automatically extract the indices; and

storing the document in the database as one of the ground truth documents if the indices are manually extracted.

2. The method of claim 1, further comprising the step of applying the document and the indices to a workflow process upon a successful automatic extraction of the indices from the document, wherein the indices provide a set of values that are processed according to the workflow process.

3. The method of claim 2, further comprising the steps of:

processing the indices according to the workflow process; and

storing the document in the database as one of the ground truth documents if the indices were successfully processed according to the workflow process.

4. The method of claim 1, wherein the step of reclassifying the document from the first one of the classifications to the second one of the classifications during the course of the automated extraction of the indices by drawing the association between the document and the at least one of the ground truth documents further comprises:

identifying a classification of the at least one of the ground truth documents; and

specifying the classification of the at least one of the ground truth documents as the second one of the classifications.

5. The method of claim 1, further comprising the steps of:

identifying a plurality of putative indices from the document; and

wherein the drawing of the association between the document and at least one of the ground truth documents is performed by searching for occurrences of the search terms in a plurality of sets of indices in the database, each set of indices being associated with a corresponding one of the ground truth documents.

6. The method of claim 1, wherein the drawing of the association between the document and at least one of the ground truth documents is performed by:

comparing a structure of the document with a structure of a plurality of the ground truth documents; and

drawing the association between the document and at least one of the ground truth documents by identifying a match between the structure of the document and the structure of the at least one of the ground truth documents.

7. A program embodied in a computer-readable medium for indexing documents, comprising:

a database that includes a plurality of ground truth documents, the ground truth documents being organized in a plurality of classifications;

code that attempts to automatically extract indices from a document based upon a classification associated with the document;

code that reclassifies the document from a first one of the classifications to a second one of the classifications during the course of the automated extraction of the indices by drawing an association between the document and at least one of the ground truth documents;

code that facilitates a manual extraction of the indices from the document upon a failure to automatically extract the indices; and

code that stores the document in the database as one of the ground truth documents if the indices are manually extracted.

8. The program embodied in the computer-readable medium of claim 7, further comprising:

a workflow process that is executed to perform a task based upon a set of values associated with the indices automatically extracted from the document; and

code that applies the indices to the workflow process upon a successful automatic extraction of the indices from the document.

9. The program embodied in the computer-readable medium of claim 8, further comprising code that stores the document and the indices in the database as one of the ground truth documents if the task associated with the workflow process was successfully performed based upon the indices.

10. The program embodied in the computer-readable medium of claim 7, wherein the code that reclassifies the document from the first one of the classifications to the second one of the classifications during the course of the automated extraction of the indices by drawing the association between the document and the at least one of the ground truth documents further comprises:

code that identifies a classification of the at least one of the ground truth documents; and

code that specifies the classification of the at least one of the ground truth documents as the second one of the classifications.

11. The program embodied in the computer-readable medium of claim 7, further comprising:

code that identifies a plurality of putative indices from the document; and

code that searches for occurrences of the putative indices in a plurality of sets of indices in the database, each set of indices being associated with a corresponding one of the ground truth documents, thereby drawing the association between the document and at least one of the ground truth documents.

12. The program embodied in the computer-readable medium of claim 7, wherein the code that reclassifies the document from the first one of the classifications to the second one of the classifications during the course of the automated extraction of the indices by drawing the association between the document and the at least one of the ground truth documents further comprises:

code that compares a structure of the document with a structure of a plurality of the ground truth documents; and

code that draws the association between the document and at least one of the ground truth documents by identifying a match between the structure of the document and the structure of the at least one of the ground truth documents.

13. An system for indexing documents, comprising:

a server having a processor and a memory;

a database stored in the memory, the database including a plurality of ground truth documents, the ground truth documents being organized in a plurality of classifications;

a document stored in the memory, the document being classified in a first one of the classifications; and

an automated document indexing system stored in the memory and executable by the processor, the automated document indexing system comprising;

logic that attempts to automatically extract indices from the document based upon the first one of the classifications;

logic that reclassifies the document from the first one of the classifications to a second one of the classifications during the course of the automated extraction of the indices by drawing an association between the document and at least one of the ground truth documents;

logic that facilitates a manual extraction of the indices from the document upon a failure to automatically extract the indices; and

logic that stores the document in the database as one of the ground truth documents if the indices are manually extracted.

14. The system of claim 13, further comprising:

a workflow processing system stored in the memory and executable by the processor, the workflow processing system being executed to perform a task based upon a set of values associated with the indices automatically extracted from the document; and

the workflow processing system further comprising logic that applies the indices to the workflow process upon a successful automatic extraction of the indices from the document.

15. The system of claim 14, wherein the workflow processing system further comprises logic that stores the document and the indices in the database as one of the ground truth documents if the task associated with the workflow process was successfully performed based upon the indices.

16. The system of claim 13, wherein the logic that reclassifies the document from the first one of the classifications to the second one of the classifications during the course of the automated extraction of the indices by drawing the association between the document and the at least one of the ground truth documents further comprises:

logic that identifies a classification of the at least one of the ground truth documents; and

logic that specifies the classification of the at least one of the ground truth documents as the second one of the classifications.

17. The system of claim 13, wherein the automated document indexing system further comprises:

logic that identifies a plurality of putative indices from the document; and

logic that searches for occurrences of the putative indices in a plurality of sets of indices in the database, each set of indices being associated with a corresponding one of the ground truth documents, thereby drawing the association between the document and at least one of the ground truth documents.

18. The system of claim 13, wherein the logic that reclassifies the document from the first one of the classifications to the second one of the classifications during the course of the automated extraction of the indices by drawing the association between the document and the at least one of the ground truth documents further comprises:

logic that compares a structure of the document with a structure of a plurality of the ground truth documents; and

logic that draws the association between the document and at least one of the ground truth documents by identifying a match between the structure of the document and the structure of the at least one of the ground truth documents.

19. An system for indexing documents, comprising:

a database stored in a memory, the database including a plurality of ground truth documents, the ground truth documents being organized in a plurality of classifications;

a document stored in the memory, the document being classified in a first one of the classifications; and

means for attempting to automatically extract indices from the document based upon the first one of the classifications;

means for reclassifying the document from the first one of the classifications to a second one of the classifications during the course of the automated extraction of the indices by drawing an association between the document and at least one of the ground truth documents;

means for facilitating a manual extraction of the indices from the document upon a failure to automatically extract the indices; and

means for storing the document in the database as one of the ground truth documents if the indices are manually extracted.

20. The system of claim 19, further comprising:

means for performing a task associated with a workflow process based upon a set of values associated with the indices automatically extracted from the document; and

means for applying the indices to the workflow process upon a successful automatic extraction of the indices from the document.

21. The system of claim 20, further comprising means for storing the document and the indices in the database as one of the ground truth documents if the task associated with the workflow process was successfully performed based upon the indices.