Methods and systems for correlating connections between users and links between articles

• US 10,402,457 B1
• Filed: 01/26/2016
• Issued: 09/03/2019
• Est. Priority Date: 12/31/2004
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google LLC (Alphabet Inc.)
• Inventors: Lawrence, Stephen R.
• Primary Examiner(s): Vanderhorst, Maria V
• Application Number:
  US15/006,887
• Time in Patent Office: 1,316 Days
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9536 Search customisation based on social or collaborative filtering

Abstract

Methods and systems for correlating connections between users and links between articles to identify search and/or ad spamming are disclosed. Social networks can be used to identify connections between users for correlation with links between articles, which can be identified through searches of article contents and/or back tracing accesses to articles. One disclosed method comprises identifying first associations between a plurality of users in a network of associated users; identifying second associations between one or more users and one or more articles; identifying third associations between at least some of the articles or between some of the users and access to some of the articles; and determining at least one of the third associations is correlated with one or more of the first associations.

Claims

1. A computer-implemented method comprising:

performing, by a search engine of a first server device, a network crawl to identify a set of linking articles;

identifying, by the search engine of the first server device, a set of first links, wherein a first link links from a linking article within the set of linking articles identified by the network crawl to a destination article within a set of destination articles;

identifying, by the search engine of the first server device, a set of second links, wherein a second link links from the destination article within the set of destination articles to the linking article within the set of linking articles identified by the network crawl;

identifying, by the search engine of the first server device, a first set of administrators, the first set of administrators being administrators of the linking articles within the set of linking articles identified by the network crawl;

identifying, by the search engine of the first server device, a second set of administrators, the second set of administrators being administrators of the destination articles within the set of destination articles;

receiving, by the search engine of the first server device, connections information, the connections information describing relationships within a social network among administrators within the first set of administrators and the second set of administrators;

correlating, by the search engine of the first server device, the connections information describing the relationships within the social network among the administrators within the first set of administrators and the second set of administrators with the first and second set of links identified as linking between articles within the set of linking articles and the set of destination articles;

determining, based on the correlation, that a first specific administrator being from the first set of administrators and associated with a first specific linking article from the set of linking articles, and a second specific administrator being from the set of second administrators and associated with a first specific destination article from the set of destination articles, share a bias toward linking to each other'"'"'s articles; and

assigning, based on the bias, searching weights to one or more of a first link linking from the first specific linking article to the first specific destination article and a second link linking from the first specific destination article to the first specific linking article, the searching weights used in determining an appearance of the set of linking articles, including the first specific linking article, and the set of destination articles, including the first specific destination article, in a search result and in reducing an effect of spamming on the search result.

2. The method of claim 1, wherein identifying the first set of administrators and the set of first links is inferred from access patterns of the first set of administrators.

3. The method of claim 1, wherein identifying the first set of administrators and the set of first links is based on identifying groups of postings where an administrator from the first set of administrators refers to the linking article.

4. The method of claim 1, further comprising ranking the search result including linking and destination articles retrieved in response to a search query based on the searching weights to reflect the bias as indicated by the correlation of the connections information with the links identified as linking between articles within the set of linking articles and the set of destination articles.

5. The method of claim 1, wherein the connections information describing the relationships within the social network among the administrators is based on a number of common entries in users'"'"' member profiles.

6. The method of claim 1, wherein the set of linking articles includes advertisements and further comprising:

identifying a first advertisement;

receiving a click-through of the first advertisement from the second specific administrator; and

discounting the click-through of the first advertisement from the second specific administrator as spamming of the first advertisement based on the bias.

7. The method of claim 1, wherein the bias is due to the relationship identified between the first specific administrator and the second specific administrator.

8. A system comprising:

one or more processors; and

a search engine stored on a memory and executable by the one or more processors of a first server device, the search engine;

performing a network crawl to identify a set of linking articles, identifying a set of first links, wherein a first link links from a linking article within the set of linking articles identified by the network crawl to a destination article within a set of destination articles,

identifying a set of second links, wherein a second link links from the destination article within the set of destination articles to the linking article within the set of linking articles identified by the network crawl,

identifying a first set of administrators, the first set of administrators being administrators of the linking articles within the set of linking articles identified by the network crawl,

identifying a second set of administrators, the second set of administrators being administrators of the destination articles within the set of destination articles,

receiving connections information, the connection information describing relationships within a social network among the first set of administrators and the second set of administrators,

correlating the connections information describing the relationships within the social network among the administrators within the first set of administrators and the second set of administrators with the first and second set of links identified as linking between articles within the set of linking articles and the set of destination articles,

determining, based on the correlation, that a first specific administrator being from the first set of administrators and associated with a first specific linking article from the set of linking articles, and a second specific administrator being from the set of second administrators and associated with a first specific destination article from the set of destination articles, share a bias toward linking to each other'"'"'s articles, and

assigning, based on the bias, searching weights to one or more of a first link linking from the first specific linking article to the first specific destination article and a second link linking from the first specific destination article to the first specific linking article, the searching weights used in determining an appearance of the set of linking articles, including the first specific linking article, and the set of destination articles, including the first specific destination article, in a search result and in reducing an effect of spamming on the search result.

9. The system of claim 8, wherein the search engine identifies the first set of administrators and the set of first links by inferring from access patterns of the first set of administrators.

10. The system of claim 8, wherein the search engine identifies the first set of administrators and the set of first links based on identifying groups of postings where an administrator from the first set of administrators refers to the linking article.

11. The system of claim 8, wherein the search engine ranks the search result including linking and destination articles retrieved in response to a search query based on the searching weights to reflect the bias as indicated by the correlation of the connections information with the links identified as linking between articles within the set of linking articles and the set of destination articles.

12. The system of claim 8, wherein the connections information describing the relationships within the social network among the administrators is based on a number of common entries in the users'"'"' member profiles.

13. The system of claim 8, wherein the set of linking articles includes advertisements and the search engine identifies a first advertisement, receives a click-through of the first advertisement from the second specific administrator, and discounts the click-through of the first advertisement from the second specific administrator as spamming of the first advertisement based on the bias.

14. The system of claim 8, wherein the bias is due to the relationship identified between the first specific administrator and the second specific administrator.

15. A computer program product comprising a non-transitory computer usable medium including a computer readable program, wherein the computer readable program when executed on a computer causes the computer to perform steps comprising:

performing, by a search engine of the computer, a network crawl to identify a set of linking articles;

identifying, by the search engine of the computer, a set of first links, wherein a first link links from a linking article within the set of linking articles identified by the network crawl to a destination article within a set of destination articles;

identifying, by the search engine of the computer, a set of second links, wherein a second link links from the destination article within the set of destination articles to the linking article within the set of linking articles identified by the network crawl;

identifying, by the search engine of the computer, a first set of administrators, the first set of administrators being administrators of the linking articles within the set of linking articles identified by the network crawl;

identifying, by the search engine of the computer, a second set of administrators, the second set of administrators being administrators of the destination articles within the set of destination articles;

receiving, by the search engine of the computer, connections information, the connections information describing relationships within a social network among administrators within the first set of administrators and the second set of administrators;

correlating, by the search engine of the computer, the connections information describing the relationships within the social network among the administrators within the first set of administrators and the second set of administrators with the first and second set of links identified as linking between articles within the set of linking articles and the set of destination articles;

determining, based on the correlation, that a first specific administrator being from the first set of administrators and associated with a first specific linking article from the set of linking articles, and a second specific administrator being from the set of second administrators and associated with a first specific destination article from the set of destination articles, share a bias toward linking to each other'"'"'s articles; and

assigning, based on the bias, searching weights to one or more of a first link linking from the first specific linking article to the first specific destination article and a second link linking from the first specific destination article to the first specific linking article, the searching weights used in determining an appearance of the set of linking articles, including the first specific linking article, and the set of destination articles, including the first specific destination article, in a search result and in reducing an effect of spamming on the search result.

16. The computer program product of claim 15, wherein identifying the first set of administrators and the set of first links is inferred from access patterns of the first set of administrators.

17. The computer program product of claim 15, wherein identifying the first set of administrators and the set of first links is based on identifying groups of postings where an administrator from the first set of administrators refers to the linking article.

18. The computer program product of claim 15, further comprising ranking the search result including linking and destination articles retrieved in response to a search query based on the searching weights to reflect the bias as indicated by the correlation of the connections information with the links identified as linking between articles within the set of linking articles and the set of destination articles.

19. The computer program product of claim 15, wherein the set of linking articles includes advertisements and further comprising:

identifying a first advertisement;

receiving a click-through of the first advertisement from the second specific administrator; and

discounting the click-through of the first advertisement from the second specific administrator as spamming of the first advertisement based on the bias.

20. The computer program product of claim 15, wherein the bias is due to the relationship identified between the first specific administrator and the second specific administrator.