Ranking of documents based on analysis of related documents

• US 8,548,995 B1
• Filed: 09/10/2003
• Issued: 10/01/2013
• Est. Priority Date: 09/10/2003
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Curtiss, Michael
• Primary Examiner(s): SYED, FARHAN M
• Application Number:
  US10/658,452
• Time in Patent Office: 3,674 Days
• Field of Search: 707/3, 707/6
• US Class Current: 707/731
• CPC Class Codes: G06F 16/338 Presentation of query results
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques

Abstract

A ranking component refines an initial document ranking in which the initial ranking is based on a scoring criterion. The ranking component generates or receive a set of documents that are related to a document. The set of documents may be related in a number of ways, such as similarity in authorship or publisher. The ranking component then applies the scoring criterion to the set of related documents to generate a related set score. The related set score is used to refine the initial rankings.

Claims

1. A computer-implemented method performed by one or more server devices, the method comprising:

identifying, using one or more processors associated with the one or more server devices, a first document;

determining, using one or more processors associated with the one or more server devices, a first relevance score for the first document,

determining the first relevance score including;

comparing a scoring criterion to the first document to determine the first relevance score;

identifying, using one or more processors associated with the one or more server devices, a set of documents that is related to the first document,

one or more documents, in the set of documents, and the first document being associated with a same web site;

calculating, using one or more processors associated with the one or more server devices, a second relevance score for the first document,

calculating the second relevance score including;

concatenating documents, in the set of documents, into a single pseudo-document, and

applying the scoring criterion to the single pseudo-document to obtain the second relevance score,

applying the scoring criterion to the single pseudo-document including;

 

comparing the scoring criterion to the single pseudo-document to obtain the second relevance score;

calculating, using one or more processors associated with the one or more server devices, a final relevance score for the first document as a weighted sum of the first relevance score and the second relevance score; and

transmitting, using a communication interface associated with the one or more server devices, a ranked list of documents to a client device,

the ranked list of documents including the first document,

the first document being ranked, within the ranked list of documents, based on the final relevance score.

2. The computer-implemented method of claim 1, where obtaining the first document, identifying the set of documents, calculating the second relevance score, and calculating the final relevance score are repeated for each of a plurality of documents.

3. The computer-implemented method of claim 1, where at least one document, in the set of documents, is related to the first document by a similarity in authorship.

4. The computer-implemented method of claim 1, where at least one document, in the set of documents, is related to the first document by a similarity in publication source.

5. The computer-implemented method of claim 1, where the one or more documents, included in the same section of the publication, include news articles.

6. The computer-implemented method of claim 5, where at least one document, in the set of documents, and the first document are included in a same section of a publication.

7. The computer-implemented method of claim 1, where the set of documents includes postings to a message board.

8. The computer-implemented method of claim 7, where at least one document, in the set of documents, and the first document are associated with a same email address of a user that submitted the postings.

9. A system comprising:

one or more server devices to;

identify a first document;

determine a first relevance score associated with the first document,

when determining the first relevance score, the one or more server devices are to;

compare a scoring criterion to the first document to determine the first relevance score;

identify a set of documents related to the first document,

when identifying the set of documents, the one or more server devices are to;

identify one or more documents, in the set of documents, based on the first document and the one or more documents being associated with a same web site;

calculate a second relevance score associated with the first document,

when calculating the second relevance score, the one or more server devices are to;

concatenate documents, in the set of documents, into a single pseudo-document, and

apply the scoring criterion to the single pseudo-document to obtain the second relevant score,

 

when applying the scoring criterion to the single pseudo-document, the one or more server devices are to;

 

compare the scoring criterion to the single pseudo-document to obtain the second relevance score;

calculate a final relevance score for the first document as a weighted sum of the first relevance score and the second relevance score; and

transmit a ranked list of documents to a client device,

the ranked list of documents including the first document,

the first document being ranked, within the ranked list of documents, based on the final relevance score.

10. The system of claim 9, where the one or more server devices are to identify the first document, identify the set of documents, calculate the second relevance score, and calculate the final relevance score for each of a plurality of documents.

11. The system of claim 9, where at least one document, in the set of documents, is related to the first document by a similarity in authorship.

12. The system of claim 9, where at least one document, in the set of documents, is related to the first document by a similarity in publication source.

13. The system of claim 9, where the first document and the set of documents include news articles.

14. The system of claim 13, where at least one document, in the set of documents, and the first document are included in a same section of a publication.

15. The system of claim 9, where the first document and at least one document, in the set of documents, include one of:

postings, of a same group, to a message board, or

postings, of a same thread, to a message board.

16. The system of claim 9, where at least one document, in the set of documents, and the first document include postings to a message group, and

where the first document and the least one document are associated with a same email address of a user that submitted the postings.

26. The system of claim 9, where the same web site is associated with:

at least two hosts that are related, or

at least two domain names that are related.

17. A non-transitory computer-readable medium storing instructions, the instructions comprising:

a plurality of instructions which, when executed by a device, cause the device to;

identify a first document;

determine a first relevance score associated with the first document,

one or more instructions, of the plurality of instructions, to determine the first relevance score including;

one or more instructions to compare a scoring criterion to the first document to determine the first relevance score;

identify a set of documents related to the first document,

one or more documents, in the set of documents, being identified based on the one or more documents and the first document being associated with a same web site;

calculate a second relevance score associated with the first document,

one or more instructions, of the plurality of instructions, to calculate the second relevance score including;

one or more instructions to concatenate documents, in the set of documents, into a single pseudo-document, and

one or more instructions to apply the scoring criterion to the single pseudo-document to obtain the second relevance score;

calculate a final relevance score for the first document,

the final relevance score being based on;

the first relevance score, and

a value that is based on;

 

the first relevance score, and

 

the second relevance score; and

transmit a ranked list of documents to a client device,

the ranked list of documents including the first document,

the first document being ranked, within the ranked list of documents, based on the final relevance score.

18. The non-transitory computer-readable medium of claim 17, where the plurality of instructions includes one or more instructions to identify the first document, identify the set of documents, calculate the second relevance score, and calculate the final relevance score, for each of a plurality of documents.

19. The non-transitory computer-readable medium of claim 17, where at least one document, in the set of documents, is identified based on a similarity in authorship between the at least one document and the first document.

20. The non-transitory computer-readable medium of claim 17, where at least one document, in the set of documents, is identified based on a similarity in publication source between the at least one document and the first document.

21. The non-transitory computer-readable medium of claim 17, where at least one document, in the set of documents, is identified based on the at least one document and the first document being:

news articles of a same news source, or

news articles of a same authorship.

22. The non-transitory computer-readable medium of claim 17, where at least one document, in the set of documents, is identified based on the first document and the at least one document being included in a same section of a publication.

23. The non-transitory computer-readable medium of claim 17, where at least one document, in the set of documents, is identified based on the at least one document and the first document being:

postings, of a same topic, to a message board, or

postings, of a same thread, to a message board.

24. The non-transitory computer-readable medium of claim 17, where at least one document, in the set of documents, and the first document are postings to a message group, and

where the at least one document is identified based on the first document and the least one document being associated with a same email address of a user associated with the postings.

25. The non-transitory computer-readable medium of claim 17, where the value is obtained by multiplying the first relevance score by the second relevance score.