Method and system for evaluating content via a computer network

• US 8,745,074 B1
• Filed: 09/14/2012
• Issued: 06/03/2014
• Est. Priority Date: 09/14/2012
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Covell, Michele, Baluja, Shumeet, Sacks, Josh, Wu, Yuchen
• Primary Examiner(s): LU, KUEN S
• Application Number:
  US13/619,467
• Time in Patent Office: 627 Days
• Field of Search: 707/758
• US Class Current: 707/758
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation

Abstract

Systems and methods of evaluating information via a computer network are provided. A content group can be identified, and each item of the content group can be associated with a vector indicating at least one user interest category of users exposed to the item. The vectors of each item can be evaluated to generate a first nearest neighbor list of each item of the content group. The nearest neighbor list of a first item can be compared with the nearest neighbor list of a second item. Based on a result of the comparison, the first and second items can be associated in a cluster.

Claims

1. A computer implemented method of evaluating information via a computer network, comprising:

identifying, by a data processing system, a first content group comprising one or more content items, each content item of the first content group having an interest category vector indicating at least one user interest category of users exposed to the content item;

evaluating, by the data processing system, the interest category vector of a content item of the first content group in conjunction with an interest category vector of a content item of each of a plurality of other content groups to calculate a plurality of first distance metrics, each first distance metric indicating a similarity between the first content group and one of the other content groups;

generating, by the data processing system, a first nearest neighbor list of the first content group, the first nearest neighbor list comprising a ranking of the other content groups based on the calculated first distance metrics;

comparing the first nearest neighbor list of the first content group with a second nearest neighbor list of a second content group to calculate a second distance metric indicating a similarity between the first nearest neighbor list and the second nearest neighbor list; and

based on the calculated second distance metric, combining the first content group with the second content group in a cluster that replaces the first and second content groups.

2. The method of claim 1, further comprising:

evaluating the interest category vector of each content item of the first content group to generate a nearest neighbor list of each content item of the first content group; and

generating the first nearest neighbor list of the first content group based at least in part on the nearest neighbor list for each content item of the first content group.

3. The method of claim 1, further comprising:

identifying a plurality of users who accessed at least one content item of the first content group;

determining user interest categories of the plurality of users; and

generating an interest category vector of the at least one item of the first content group accessed by the plurality of users based on an evaluation of the user interest categories of the plurality of users.

4. The method of claim 1, further comprising:

determining an edit distance between the first nearest neighbor list of the first content group and the second nearest neighbor list of the second content group; and

associating the first content group with the second content group in the cluster based on the edit distance.

5. The method of claim 1, further comprising:

evaluating the interest category vector of each content item of the first content group to determine at least one degree of similarity between each content item of the first content group.

6. The method of claim 5, wherein the first nearest neighbor list orders a first content item of the first content group relative to a second content item of the first content group based on the degree of similarity of the first content item with the second content item.

7. The method of claim 1, further comprising:

identifying a group of users as having accessed a first content item of the first content group;

obtaining a plurality of user interest categories of the group of users; and

determining a first interest category vector of the first content item based on the plurality of user interest categories.

8. The method of claim 7, further comprising at least one of:

determining that a user of the group of users clicked on the first content item; and

determining that the first content item was rendered for display on a computing device associated with the user of the group of users.

9. The method of claim 1, further comprising:

comparing the first nearest neighbor list with a third nearest neighbor list of a third content group;

based on a result of the comparison, excluding the third content group from the cluster.

10. A system of evaluating information via a computer network, comprising:

a data processing system having at least one of a memory unit configured to store processor executable instructions and a processor configured to execute the processor executable instructions to;

identify a first content group comprising one or more content items, each content item of the first content group having an interest category vector indicating at least one user interest category of users exposed to the content item;

evaluate the interest category vector of a content item of the first content group in conjunction with an interest category vector of a content item of each of a plurality of other content groups to calculate a plurality of first distance metrics, each first distance metric indicating a similarity between the first content group and one of the other content groups;

generate a first nearest neighbor list of the first content group, the first nearest neighbor list comprising a ranking of the other content groups based on the calculated first distance metrics;

compare the first nearest neighbor list of the first content group with a second nearest neighbor list of a second content group to calculate a second distance metric indicating a similarity between the first nearest neighbor list and the second nearest neighbor list; and

combine the first content group with the second content group in a cluster that replaces the first and second content groups based on the calculated second distance metric.

11. The system of claim 10, wherein the data processing system is further configured to:

identify a plurality of users who accessed at least one content item of the first content group;

determine user interest categories of the plurality of users; and

generate an interest category vector of the at least one content item of the first content group accessed by the plurality of users based on an evaluation of the user interest categories of the plurality of users.

12. The system of claim 10, wherein the data processing system further configured to:

determine an edit distance between the first nearest neighbor list of the first content group and the second nearest neighbor list of the second content group; and

associate the first content group with the second content group based on the edit distance.

13. The system of claim 10, wherein the data processing system is further configured to:

evaluate the interest category vector of each content item of the first content group to determine at least one degree of similarity between each content item of the first content group.

14. The system of claim 13, wherein the first nearest neighbor list orders a first content item of the first content group relative to a second content item of the first content group based on the degree of similarity of the first content item with the second content item.

15. The system of claim 10, wherein the data processing system is further configured to:

identify a group of users as having accessed a first content item of the first content group;

obtain a plurality of user interest categories of the group of users; and

determine a first interest category vector of the first content item based on the plurality of user interest categories.

16. The system of claim 15, further comprising at least one of:

the data processing system configured to determine that a user of the group of users clicked on the first content item; and

the data processing system configured to determine that the first content item was rendered for display on a computing device associated with the user of the group of users.

17. The system of claim 10, wherein the data processing system is further configured to:

compare the first nearest neighbor list with a third nearest neighbor list of a third content group;

exclude the third content group from the cluster based on a result of the comparison.

18. A computer readable storage medium having instructions encoded thereon that when executed by at least one processor cause the at least one processor to perform a method to evaluate information obtained via a computer network, the method comprising:

identifying a first content group comprising one or more content items, each content item of the first content group having an interest category vector indicating at least one user interest category of users exposed to the content item;

evaluating the interest category vector of a content item of the first content group in conjunction with an interest category vector of a content item of each of a plurality of other content groups to calculate a plurality of first distance metrics, each first distance metric indicating a similarity between the first content group and one of the other content groups;

generating a first nearest neighbor list of the first content group, the first nearest neighbor list comprising a ranking of the other content groups based on the calculated first distance metrics;

comparing the first nearest neighbor list of the first content group with a second nearest neighbor list of a second content group to calculate a second distance metric indicating a similarity between the first nearest neighbor list and the second nearest neighbor list; and

combining the first content group with the second content group in a cluster that replaces the first and second content groups based on the calculated second distance metric.

19. The computer readable storage medium of claim 18, wherein the method further comprises:

identifying a plurality of users who accessed at least one content item of the first content group;

determining user interest categories of the plurality of users; and

generating an interest category vector of the at least one content item of the first content group accessed by the plurality of users based on an evaluation of the user interest categories of the plurality of users.

20. The computer readable storage medium of claim 18, wherein the method further comprises:

evaluating the interest category vector of each content item of the first content group to determine at least one degree of similarity between each content item of the first content group.