Page-ranking via user expertise and content relevance

• US 7,676,464 B2
• Filed: 03/17/2006
• Issued: 03/09/2010
• Est. Priority Date: 03/17/2006
• Status: Active Grant

• Current Assignee: X Corp. (f/k/a Twitter, Inc.) (X Holdings Corp.)
• Original Assignee: International Business Machines Corporation
• Inventors: Milch, Daniel, Laker, Meir M., Lenchner, Jonathan
• Primary Examiner(s): Vital; Pierre M
• Assistant Examiner(s): LIAO, JASON G
• Application Number:
  US11/377,413
• Publication Number:
  US 20070219993A1
• Time in Patent Office: 1,453 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access

Abstract

A page-ranking method includes mining a portion of content of a user workstation which is connectable to a network to detect references to pages of the network. The pages may be ranked based on the detected references.

Claims

1. A page-ranking method, comprising:

designating a portion of content stored on a hard drive of plural user workstations which are connectable to a network, such that said portion is not accessible for mining to detect references to plural pages on said network;

developing a set of information categories which are relevant to said network;

assigning a content relevance rank to said plural pages for said set of categories, said content relevance rank indicating a degree of relevance of a page of said relevance rank indicating a degree of relevance of a page of said plural pages in a category of said plural categories;

assigning an expertise rank to plural persons associated with said plural user work stations in said categories, said expertise ranking indicating a level of expertise for a person of said plural persons with respect to a category of said plural categories; and

computing a value rank for said plural pages in said plural categories, said value rank indicating a level of value for a page of said plural pages with respect to a category of said plural categories, said computing said value rank comprising;

mining a portion of said content stored on said hard drive of said plural user workstations which is other than said designated portion to detect references to said plural pages, and counting a number of said references to said plural pages; and

if a reference to a page is detected in a mined portion of said content of a user workstation, then increasing said value rank of said referenced page in said plural categories based on the content relevance rank of the referenced page and the expertise rank of a person of said plural persons who is associated with said user workstation,

if a reference to a page is detected in an e-mail in the mined portion of the content of the user workstation, then increasing the value rank of the referenced page based on the content relevance rank of the referenced page, the expertise rank of the sender of the e-mail and the expertise rank of the recipient of the e-mail

wherein if a reference to a page is detected in an e-mail, then said value rank of said page is adjusted based on the expertise rank of the sender and a recipient of said page, and the relevance rank of the page, and if a reference to a page is detected in a message passed through an instance messaging facility, then said value rank of said page is adjusted based on an expertise rank of a sender of said message, an expertise rank of a recipient of said message, and the relevance rank of the page

wherein said mined portion of content comprises a file that is textually minable and is one of residing on and attached to a user workstation,

wherein said network comprises an intranet, and mining said portion of said content is performed as part of a back-up process for said intranet,

wherein said mining comprises using at least one of scanning code and crawling code to detect said references,

wherein said assigning said expertise rank comprises assigning a weight to a user workstation based on references detected in said mined portion of content, and based on a time at which high quality pages are stored on said user workstation,

wherein said mined portion of content comprises browser bookmarks, e-mails, instant messages and unstructured text files, and

wherein said mining said portion of said content, and ranking said pages are performed routinely and periodically.

2. The method of claim 1, wherein said mining said portion of content comprises using a software implemented mining device which is stored as software on a central server of said network.

3. The method of claim 1, wherein said mining said portion of content is scheduled independently by said user of said user workstation.

4. The method of claim 1, wherein said mining said portion of content is scheduled to be performed at a time when the user workstation is likely to be idle.

5. The method of claim 1, wherein said reference to said page comprises a hyperlink to said page.

19. The method of claim 1, wherein said assigning said expertise rank comprises:

assigning an initial expertise rank for said person in a category of said categories; and

increasing said initial expertise rank based on a number of references detected in a workstation of said person in said category.

20. The method of claim 1, wherein said page on said network comprises an intranet document.

6. An intranet search method, comprising:

entering a search query;

ranking plural pages of said network, comprising

designating a portion of content stored on a hard drive of plural user workstations which are connectable to a network, such that said designated portion is not accessible for mining to detect references to plural pages on said network;

developing a set of information categories which are relevant to said network;

assigning a content relevance rank to said plural pages for said set of categories, said content relevance rank indicating a degree of relevance of a page of said relevance rank indicating a degree of relevance of a page of said plural pages in a category of said plural categories;

assigning an expertise rank to plural persons associated with said plural user work stations in said categories, said expertise ranking indicating a level of expertise for a person of said plural persons with respect to a category of said plural categories; and

computing a value rank for said plural pages in said plural categories, said value rank indicating a level of value for a page of said plural pages with respect to a category of said plural categories, said computing said value rank comprising;

mining a portion of said content stored on said hard drive of said plural user workstations which is other than said designated portion to detect references to said plural pages, and counting a number of said references to said plural pages; and

if a reference to a page is detected in a mined portion of said content of a user workstation, then increasing said value rank of said referenced page in said plural categories based on the content relevance rank of the referenced page and the expertise rank of a person of said plural persons who is associated with said user workstation,

if a reference to a page is detected in an e-mail in the mined portion of the content of the user workstation, then increasing the value rank of the referenced page based on the content relevance rank of the referenced page, the expertise rank of the sender of the e-mail and the expertise rank of the recipient of the e-mail

wherein if a reference to a page is detected in an e-mail, then said value rank of said page is adjusted based on the expertise rank of the sender and a recipient of said page, and the relevance rank of the page, and if a reference to a page is detected in a message passed through an instance messaging facility, then said value rank of said page is adjusted based on an expertise rank of a sender of said message, an expertise rank of a recipient of said message, and the relevance rank of the page

wherein said mined portion of content comprises a file that is textually minable and is one of residing on and attached to a user workstation,

wherein said network comprises an intranet, and mining said portion of said content is performed as part of a back-up process for said intranet,

wherein said mining comprises using at least one of scanning code and crawling code to detect said references,

wherein said assigning said expertise rank comprises assigning a weight to a user workstation based on references detected in said mined portion of content, and based on a time at which high quality pages are stored on said user workstation,

wherein said mined portion of content comprises browser bookmarks, e-mails, instant messages and unstructured text files, and

wherein said mining said portion of said content, and ranking said pages are performed routinely and periodically.

7. The method of claim 6, wherein said returning said list of pages comprises returning said list of pages in an order which is a function of the pages'"'"' value ranks.

8. The method of claim 6, wherein said returning said list of pages comprises returning said list of pages in an order which is based on the sum of the pages'"'"' value ranks across all of the categories.

9. The method of claim 6, wherein said returning said list of pages comprises returning said list of pages in an order which is based on the pages'"'"' highest value rank in any category.

10. The method of claim 6, wherein said returning said list of pages comprises returning said list of pages in an order which is based on the pages'"'"' highest median value rank across the categories.

11. The method of claim 6, wherein a keyword search term used in said search may be categorized into one of the relevant categories in the value ranking, and said returning said list of pages comprises returning said list of pages in an order which is based on the value ranks of the page in the category into which the search term is categorized.

12. An page-ranking method, comprising:

designating a portion of content stored on a hard drive of plural user workstations which are connectable to a network, such that said designated portion is not accessible for mining to detect references to plural pages on said network;

developing a set of information categories which are relevant to said network;

assigning a content relevance rank to said plural pages for said set of categories, said content relevance rank indicating a degree of relevance of a page of said relevance rank indicating a degree of relevance of a page of said plural pages in a category of said plural categories;

assigning an expertise rank to plural persons associated with said plural user work stations in said categories, said expertise ranking indicating a level of expertise for a person of said plural persons with respect to a category of said plural categories; and

computing a value rank for said plural pages in said plural categories, said value rank indicating a level of value for a page of said plural pages with respect to a category of said plural categories; and

computing a value rank for said plural pages in said plural categories, said value rank indicating a level of value for a page of said plural pages with respect to a category of said plural categories, said computing said value rank comprising;

mining a portion of said content stored on said hard drive of said plural user workstations which is other than said designated portion to detect references to said plural pages, and counting a number of said references to said plural pages; and

if a reference to a page is detected in a mined portion of said content of a user workstation, then increasing said value rank of said referenced page in said plural categories based on the content relevance rank of the referenced page and the expertise rank of a person of said plural persons who is associated with said user workstation,

if a reference to a page is detected in an e-mail in the mined portion of the content of the user workstation, then increasing the value rank of the referenced page based on the content relevance rank of the referenced page, the expertise rank of the sender of the e-mail and the expertise rank of the recipient of the e-mail.

13. The method of claim 12, wherein if a reference to a page is detected in an e-mail, then said value rank of said page is adjusted based on the expertise rank of a sender and a recipient of said page, and the relevance rank of the page, and if a reference to a page is detected in a message passed through an instant messaging facility, then said value rank of said page is adjusted based on an expertise rank of a sender of said message, an expertise rank of a recipient of said message, and the relevance rank of the page.

14. The method of claim 13, wherein said mined portion of content comprises a file that is textually minable and is one of residing on and attached to a user workstation.

15. The method of claim 14, said network comprises an intranet, and mining said portion of said content is performed as part of a back-up process for said intranet, and

wherein said mining comprises using at least one of scanning code and crawling code to detect said references.

16. The method of claim 15, wherein said assigning said expertise rank comprises assigning a weight to a user workstation based on references detected in said mined portion of content, and based on a time at which high quality pages are stored on said user workstation.

17. The method of claim 16, wherein said mined portion of content comprises browser bookmarks, e-mails, instant messages and unstructured text files.

18. The method of claim 17, wherein said mining said portion of said content, and ranking said pages are performed routinely and periodically.