Global anchor text processing

• US 7,788,253 B2
• Filed: 12/28/2006
• Issued: 08/31/2010
• Est. Priority Date: 12/28/2006
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Lempel, Ronny, Grennan, Frederick Stephen, Neumann, Andreas
• Primary Examiner(s): Cottingham; John R
• Assistant Examiner(s): Reyes; Mariela D
• Application Number:
  US11/617,401
• Publication Number:
  US 20080162425A1
• Time in Patent Office: 1,342 Days
• Field of Search: 707/709, 707/711
• US Class Current: 707/709
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 40/134 Hyperlinking

Abstract

Provided are techniques for building a search index. While building the search index and using the search index to respond to one or more search requests, an anchor information store is maintained, wherein each entry of the anchor information store identifies a referring document, a target document, and anchor text associated with a link from the referring document to the target document; a document is received for processing; one or more entries in the anchor information store for which the document to be processed is identified as the target document are located; anchor text is retrieved from each of the identified entries; and the retrieved anchor text is stored in an entry of the search index for the document.

Claims

1. A method for building a search index:

while building the search index and using the search index to respond to one or more search requests and performing synchronous anchor text processing,

maintaining an anchor information store, wherein each entry of the anchor information store identifies a referring document, a target document, and anchor text associated with a link from the referring document to the target document;

maintaining a rebuild agenda, wherein each entry of the rebuild agenda identifies a target document that has an entry in the search index and whose anchor text is to be updated in the search index with asynchronous processing because there is at least one new or updated link pointing to the target document;

receiving a document for processing;

for each outgoing link in the document that points to a target document,

adding an entry to the anchor information store that identifies the received document, the target document, and anchor text; and

adding an entry to the rebuild agenda for the target document; and

for each link pointing from a referring document to the document,

locating one or more entries in the anchor information store for which the received document to be processed is identified as the target document;

retrieving anchor text from each of the identified entries; and

storing the retrieved anchor text in an entry of the search index for the received document; and

performing asynchronous anchor text processing to incrementally update current entries in the search index for each document identified in each entry in the rebuild agenda in parallel with the building of the search index and in parallel with using the search index to respond to one or more search requests by;

selecting a first target document in the rebuild agenda;

using the anchor information store to find anchor text for the first target document by identifying one or more entries in the anchor information store for which the first target document is identified as the target document in the anchor information store;

retrieving anchor text from each of the identified entries; and

incrementally updating the anchor text in the entry of the search index for the first target document.

2. The computer-implemented method of claim 1, wherein the received document in the entry added to the anchor information store for the outgoing link is a referring document.

3. The computer-implemented method of claim 1, further comprising:

when a document is one of missing and unavailable,

locating one or more entries in the anchor information store for which the document is a referring document;

for each of the one or more entries, registering the target document of that entry in the rebuild agenda; and

deleting the one or more entries from the anchor information store.

4. The computer-implemented method of claim 1, further comprising:

when a document is one of missing and unavailable,

locating one or more entries in the anchor information store for which the document is the target document; and

reporting the referring document in each of the identified entries to one of a search engine administrator and a content owner of the referring document.

5. The computer-implemented method of claim 1, further comprising:

generating a graph that shows connections of nodes using the anchor information store to enable studying site connectedness.

6. The computer-implemented method of claim 1, further comprising:

adjusting a search-result weight associated with the document based on an in-link count determined by counting a number of referring documents for the document in the anchor information store.

7. The computer-implemented method of claim 1, wherein the anchor information store includes tags, which are user annotations of documents.

8. The computer-implemented method of claim 1, further comprising:

when the document has been updated with new content,

comparing existing entries in the anchor information store for which the document is a referring document with a current set of anchors in the document;

deleting each entry from the anchor information store that no longer matches the document content;

adding a new entry to the anchor information store for each new anchor in the document; and

registering a target document of any entry that was one of deleted from and added to the anchor information store in the rebuild agenda.

9. A computer program product comprising a computer readable storage medium storing a computer readable program, wherein the computer readable program when executed by a processor on a computer causes the computer to:

while building the search index and using the search index to respond to one or more search requests and performing synchronous anchor text processing,

maintain an anchor information store, wherein each entry of the anchor information store identifies a referring document, a target document, and anchor text associated with a link from the referring document to the target document;

maintain a rebuild agenda, wherein each entry of the rebuild agenda identifies a target document that has an entry in the search index and whose anchor text is to be updated in the search index with asynchronous processing because there is at least one new or updated link pointing to the target document;

receive a document for processing;

for each outgoing link in the document that points to a target document,

add an entry to the anchor information store that identifies the received document, the target document, and anchor text; and

add an entry to the rebuild agenda for the target document; and

for each link pointing from a referring document to the document, locate one or more entries in the anchor information store for which the received document to be processed is identified as the target document;

retrieve anchor text from each of the identified entries; and

store the retrieved anchor text in an entry of the search index for the received document; and

perform asynchronous anchor text processing to incrementally update current entries in the search index for each document identified in each entry in the rebuild agenda in parallel with the building of the search index and in parallel with using the search index to respond to one or more search requests by;

selecting a first target document in the rebuild agenda;

using the anchor information store to find anchor text for the first target document by identifying one or more entries in the anchor information store for which the first target document is identified as the target document in the anchor information store;

retrieving anchor text from each of the identified entries; and

updating the anchor text in the entry of the search index for the first target document.

10. The computer program product of claim 9, wherein the received document in the entry added to the anchor information store for the outgoing link is a referring document.

11. The computer program product of claim 9, wherein the computer readable program when executed on a computer causes the computer to:

when a document is one of missing and unavailable,

locate one or more entries in the anchor information store for which the document is a referring document;

for each of the one or more entries, register the target document of that entry in the rebuild agenda; and

delete the one or more entries from the anchor information store.

12. The computer program product of claim 9, wherein the computer readable program when executed on a computer causes the computer to:

when a document is one of missing and unavailable,

locate one or more entries in the anchor information store for which the document is the target document; and

report the referring document in each of the identified entries to one of a search engine administrator and a content owner of the referring document.

13. The computer program product of claim 9, wherein the computer readable program when executed on a computer causes the computer to:

generate a graph that shows connections of nodes using the anchor information store to enable studying site connectedness.

14. The computer program product of claim 9, wherein the computer readable program when executed on a computer causes the computer to:

adjust a search-result weight associated with the document based on an in-link count determined by counting a number of referring documents for the document in the anchor information store.

15. The computer program product of claim 9, wherein the anchor information store includes tags, which are user annotations of documents.

16. The computer program product of claim 9, wherein the computer readable program when executed on a computer causes the computer to:

when the document has been updated with new content,

compare existing entries in the anchor information store for which the document is a referring document with a current set of anchors in the document;

delete each entry from the anchor information store that no longer matches the document content;

add a new entry to the anchor information store for each new anchor in the document; and

register a target document of any entry that was one of deleted from and added to the anchor information store in the rebuild agenda.

17. A system for building a search index, comprising:

hardware logic capable of performing operations, the operations comprising;

while building the search index and using the search index to respond to one or more search requests and performing synchronous anchor text processing;

maintaining an anchor information store, wherein each entry of the anchor information store identifies a referring document, a target document, and anchor text associated with a link from the referring document to the target document;

maintaining a rebuild agenda, wherein each entry of the rebuild agenda identifies a target document that has an entry in the search index and whose anchor text is to be updated in the search index with asynchronous processing because there is at least one new or updated link pointing to the target document;

receiving a document for processing;

for each outgoing link in the document that points to a target document,

adding an entry to the anchor information store that identifies the received document, the target document, and anchor text; and

adding an entry to the rebuild agenda for the target document; and

for each link pointing from a referring document to the document, locating one or more entries in the anchor information store for which the received document to be processed is identified as the target document;

retrieving anchor text from each of the identified entries; and

storing the retrieved anchor text in an entry of the search index for the received document; and

performing asynchronous anchor text processing to incrementally update current entries in the search index for each document identified in each entry in the rebuild agenda in parallel with the building of the search index and in parallel with using the search index to respond to one or more search requests by;

selecting a first target document in the rebuild agenda;

using the anchor information store to find anchor text for the first target document by identifying one or more entries in the anchor information store for which the first target document is identified as the target document in the anchor information store;

retrieving anchor text from each of the identified entries; and

updating the anchor text in the entry of the search index for the first target document.

18. The system of claim 17, wherein the received document in the entry added to the anchor information store for the outgoing link is a referring document.

19. The system of claim 17, wherein the operations further comprise:

when a document is one of missing and unavailable,

locating one or more entries in the anchor information store for which the document is a referring document;

for each of the one or more entries, registering the target document of that entry in the rebuild agenda; and

deleting the one or more entries from the anchor information store.

20. The system of claim 17, wherein the operations further comprise:

when a document is missing or unavailable,

locating one or more entries in the anchor information store for which the document is the target document; and

reporting the referring document in each of the identified entries to one of a search engine administrator and a content owner of the referring document.

21. The system of claim 17, wherein the operations further comprise:

generating a graph that shows connections of nodes using the anchor information store to enable studying site connectedness.

22. The system of claim 17, wherein the operations further comprise:

adjusting a search-result weight associated with the document based on an in-link count determined by counting a number of referring documents for the document in the anchor information store.

23. The system of claim 17, wherein the anchor information store includes tags, which are user annotations of documents.

24. The system of claim 17, wherein the operations further comprise:

when the document has been updated with new content,

comparing existing entries in the anchor information store for which the document is a referring document with a current set of anchors in the document;

deleting each entry from the anchor information store that no longer matches the document content;

adding a new entry to the anchor information store for each new anchor in the document; and

registering a target document of any entry that was one of deleted from and added to the anchor information store in a rebuild agenda.