Method and apparatus for determining relative relevance between portions of large electronic documents

• US 20050154761A1
• Filed: 01/12/2004
• Published: 07/14/2005
• Est. Priority Date: 01/12/2004
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Wang, Ningning, Matada, Anmol Neelammna, Lee, Chris Guo
• Granted Patent: US 7,254,587 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/907 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: Y10S 707/99931 Database or file accessing
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing
• CPC Class Codes: Y10S 707/99948 Application of database or data structure, e.g. distributed, multimedia, or image

Abstract

A method and apparatus for determining the relative relevance of electronic documents based on metadata/content associated with the document as a whole and/or metadata/content associated with one or more subdivisions of the electronic document are provided. With the method and apparatus, metadata is associated with the document and various subdivision markers in the code of the document. A comparison of electronic documents may be made by comparing the metadata/content associated with the document and/or the subdivisions of the document to determine which documents contains subject matter that is relevant to the subject matter of another document or search criteria. The metadata/content associated with the document and/or subdivisions may be weighted and these weights may be modified based on a rank profile that may be established by a relative relevance search engine provider or may be customized by users to their specific needs. In this way, a relevance score may be determined based on the comparison of the metadata/content for the documents and/or subdivisions of the documents as well as the weights attributed to the various subdivisions and documents.

Claims

1. A method, in a data processing system, of identifying relevant documents to a portion of a first document, comprising:

receiving an identification of a portion of a first document from a client device;

identifying metadata associated with the portion of the first document;

retrieving metadata for a portion of a second document;

comparing the metadata associated with the portion of the first document to the metadata for the portion of the second document; and

determining a relative relevance of the portion of the second document to the portion of the first document based on the comparison.

2. The method of claim 1, wherein determining a relative relevance of the portion of the second document to the portion of the first document includes:

generating a correspondence value based on the comparison of the metadata associated with the portion of the first document to the metadata associated with the portion of the second document; and

generating a score value for the portion of the second document based on the correspondence value for the portion of the second document.

3. The method of claim 2, further comprising:

ranking the second document relative to other documents that have been compared to the first document based on the generated score value to obtain a ranked list of documents; and

providing the ranked list to the client device.

4. The method of claim 2, further comprising:

comparing content of the portion of the first document to content of the portion of the second document; and

generating a content value for the portion of the second document based on the comparison of the content, wherein the correspondence value and the content value for the portion of the second document are combined to obtain the generated score value.

5. The method of claim 1, wherein the portion of the first document is one of the entire first document, a chapter of the first document, a paragraph of the first document, a subdivision of the first document, and a search term.

6. The method of claim 1, wherein the portion of the second document is at least one of an entire portion of the one or more second documents, a chapter of the one or more second documents, a paragraph of the one or more second documents, and a subdivision of the one or more second documents.

7. The method of claim 1, wherein one or more first weights are associated with the portion of the first document, one or more second weights are associated with the portion of the second document, and wherein generating a score value for the portion of the second document based on the correspondence value for the portion of the second document includes applying the one or more first weights and the one or more second weights to the correspondence value.

8. The method of claim 4, wherein one or more first weights are associated with the portion of the first document, one or more second weights are associated with the portion of the second document, and wherein generating a score value for the portion of the second document based on the correspondence value and the content value for the portion of the second document includes applying the one or more first weights and the one or more second weights to the correspondence value and the content value.

9. The method of claim 8, wherein the score value is calculated using the following equation:

Comp(D_a,D_b)=W_doc-meta(a)*W_doc-meta(b)*Comp(m_a-doc, m_b-doc)+W_doc-content(a)*W_doc-content(b)*Comp(c_a-doc,c_b-doc)

where Comp(D_a,D_b) is a comparison value for a comparison of a portion of a first document D_a and a second document D_b, W_doc-meta(a) represents one or more first weights associated with metadata of the portion of the first document, W_doc-meta(b) represents one or more second weights associated with metadata of the portion of the second document, Comp(m_a-doc, m_b-doc) is a comparison value associated with a comparison of the metadata for a portion of a first document m_a-doc and the metadata for a portion of a second document m_b-doc, W_doc-content(a) represents one or more first weights associated with content of the portion of the first document, W_doc-content(b) represents one or more second weights associated with content of the portion of the second document, and Comp(c_a-doc,c_b-doc) is a comparison value obtained from comparing the content of the portion of the first document c_a-doc to the content of the portion of the second document c_b-doc.

10. The method of claim 8, further comprising:

retrieving one or more weight modifiers from a profile; and

applying the one or more weight modifiers to one or more of the first weights and the second weights.

11. A computer program product in a computer readable medium for identifying documents relevant to a portion of a first document, comprising:

first instructions for receiving an identification of a portion of a first document from a client device;

second instructions for identifying metadata associated with the portion of the first document;

third instructions for retrieving metadata for a portion of a second document;

fourth instructions for comparing the metadata associated with the portion of the first document to the metadata for the portion of the second document; and

fifth instructions for determining a relative relevance of the portion of the second document to the portion of the first document based on the comparison.

12. The computer program product of claim 11, wherein the fifth instructions for determining a relative relevance of the portion of the second document to the portion of the first document include:

instructions for generating a correspondence value based on the comparison of the metadata associated with the portion of the first document to the metadata associated with the portion of the second document; and

instructions for generating a score value for the portion of the second document based on the correspondence value for the portion of the second document.

13. The computer program product of claim 12, further comprising:

instructions for ranking the second document relative to other documents that have been compared to the first document based on the generated score value to obtain a ranked list of documents; and

instructions for providing the ranked list to the client device.

14. The computer program product of claim 12, further comprising:

instructions for comparing content of the portion of the first document to content of the portion of the second document; and

instructions for generating a content value for the portion of the second document based on the comparison of the content, wherein the correspondence value and the content value for the portion of the second document are combined to obtain the generated score value.

15. The computer program product of claim 11, wherein the portion of the first document is one of the entire first document, a chapter of the first document, a paragraph of the first document, a subdivision of the first document, and a search term.

16. The computer program product of claim 11, wherein the portion of the second document is at least one of an entire portion of the one or more second documents, a chapter of the one or more second documents, a paragraph of the one or more second documents, and a subdivision of the one or more second documents.

17. The computer program product of claim 11, wherein one or more first weights are associated with the portion of the first document, one or more second weights are associated with the portion of the second document, and wherein the instructions for generating a score value for the portion of the second document based on the correspondence value for the portion of the second document include instructions for applying the one or more first weights and the one or more second weights to the correspondence value.

18. The computer program product of claim 14, wherein one or more first weights are associated with the portion of the first document, one or more second weights are associated with the portion of the second document, and wherein the instructions for generating a score value for the portion of the second document based on the correspondence value and the content value for the portion of the second document include instructions for applying the one or more first weights and the one or more second weights to the correspondence value and the content value.

19. The computer program product of claim 18, further comprising:

instructions for retrieving one or more weight modifiers from a profile; and

instructions for applying the one or more weight modifiers to one or more of the first weights and the second weights.

20. An apparatus for identifying documents relevant to a portion of a first document, comprising:

means for receiving an identification of a portion of a first document from a client device;

means for identifying metadata associated with the portion of the first document;

means for retrieving metadata for a portion of a second document;

means for comparing the metadata associated with the portion of the first document to the metadata for the portion of the second document; and

means for determining a relative relevance of the portion of the second document to the portion of the first document based on the comparison.