Methods and systems for improving a search ranking using article information

• US 10,423,679 B2
• Filed: 02/10/2015
• Issued: 09/24/2019
• Est. Priority Date: 12/31/2003
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google LLC (Alphabet Inc.)
• Inventors: Khan, Omar Habib, Lawrence, Stephen
• Primary Examiner(s): Trujillo, James
• Assistant Examiner(s): Khoshnoodi, Fariborz
• Application Number:
  US14/617,931
• Publication Number:
  US 20190155951A1
• Time in Patent Office: 1,687 Days
• Field of Search: 707722, 707723
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation

Abstract

Systems and methods that improve client-side searching are described. In one aspect, a system and method for identifying an article, and determining a ranking score for the article based at least in part on duration data, access data, URL data, or trajectory data associated with the article is described.

Claims

1. A method performed by at least one computer processor, the method comprising:

receiving an implicit search query generated in response to client-side behavior data associated with a user, the receiving the implicit search query comprising;

receiving, from a user device, monitored user inputs of the user into an application environment that;

(i) is different from a search web page hosted by a search system, and (ii) is not an environment of a search system, wherein a monitoring of the monitored user inputs is not initiated by the user;

determining, based on the monitored user inputs and the client-side behavior data associated with the user, whether a phrase of a plurality of phrases, that includes a first set of terms and input by the user inputs into the application environment, implicitly defines a search query, wherein each phrase that implicitly defines a search query specifies the user is searching for a particular item, wherein only a subset of phrases implicitly defines a search query; and

for each phrase in the subset of phrases that implicitly defines a search query;

generating an implicit search query that includes a second set of terms that are different from the first set of terms and derived from the first set of terms;

identifying an article relevant to the received implicit search query and the particular item;

determining a ranking score for the article based at least in part on the client-side behavior data;

arranging the identified article with at least one other article in a ranking order responsive at least in part to the ranking score; and

providing, to the user device, data for presentation of the arranged articles to the user device in the ranking order.

2. The method of claim 1, wherein the identified article is a client-side article or a network article.

3. The method of claim 1, wherein the client-side behavior data further includes one or more of duration data, access data, URL data, or trajectory data.

4. The method of claim 1, wherein identifying the article comprises identifying a frequently accessed article for a one-click-away page.

5. The method of claim 3, wherein the duration data comprises determining a graph, wherein the graph comprises nodes representing URLs and edges representing connections between the nodes.

6. The method of claim 3, wherein the duration data comprises an amount of time spent on a request path.

7. The method of claim 3, wherein the duration data comprises an amount of time spent on articles within a specified number of links of a particular article.

8. The method of claim 3, further comprising:

outputting a search results listing;

determining duration data for results within the search results listing; and

associating results with a shorter duration with a lower ranking score.

9. The method of claim 3, wherein the duration data distinguishes between durations when the user is actively using a client application, and when the user is idle or performing an activity with a different client application.

19. The method of claim 1, wherein the client-side behavior data includes, at least in part, trajectory data indicating whether articles accessed by the user are web pages that started the user on a sequence of associated web pages.

10. A computer-readable medium storing computer-executable instructions which, when executed, cause at least one computer processor to perform operations comprising:

receiving an implicit search query generated in response to client-side behavior data associated with a user, the receiving the implicit search query comprising;

receiving, from a user device, monitored user inputs of the user into an application environment that;

(i) is different from a search web page hosted by a search system, and (ii) is not an environment of a search system, wherein a monitoring of the monitored user inputs are not initiated by the user;

determining, based on the monitored user inputs and the client-side behavior data associated with the user, whether a phrase of a plurality of phrases, that includes a first set of terms and input by the user inputs into the application environment, implicitly defines a search query, wherein each phrase that implicitly defines a search query specifies the user is searching for a particular item, wherein only a subset of phrases implicitly define a search query; and

for each phrase in the subset of phrases that implicitly defines a search query;

generating an implicit search query that includes a second set of terms that are different from the first set of terms and derived from the first set of terms;

identifying an article relevant to the received implicit search query and the particular item;

determining a ranking score for the article based at least in part on the client-side behavior data;

arranging the identified article with at least one other article in a ranking order responsive at least in part to the ranking score; and

providing, to the user device, data for presentation of the arranged articles to the user device in the ranking order.

11. The computer-readable medium of claim 10, wherein the identified article is a client-side article or a network article.

12. The computer-readable medium of claim 10, wherein the client-side behavior data further includes one or more of duration data, access data, URL data, or trajectory data.

13. The computer-readable medium of claim 10, wherein identifying the article comprises identifying a frequently accessed article for a one-click-away page.

14. The computer-readable medium of claim 12, wherein the duration data comprises determining a graph, wherein the graph comprises nodes representing URLs and edges representing connections between the nodes.

15. The computer-readable medium of claim 12, wherein the duration data comprises an amount of time spent on a request path.

16. The computer-readable medium of claim 12, wherein the duration data comprises an amount of time spent on articles within a specified number of links of a particular article.

17. The computer-readable medium of claim 12, wherein the instructions further cause the at least one computer processor to perform operations comprising:

outputting a search results listing;

determining duration data for results within the search results listing; and

associating results with a shorter duration with a lower ranking score.

18. The computer-readable medium of claim 12, wherein the duration data distinguishes between durations when the user is actively using a client application, and when the user is idle or performing an activity with a different client application.

20. The computer-readable medium of claim 10, wherein the client-side behavior data includes, at least in part, trajectory data indicating whether articles accessed by the user are web pages that started the user on a sequence of associated web pages.