Preferred sites

• US 9,317,563 B1
• Filed: 10/23/2013
• Issued: 04/19/2016
• Est. Priority Date: 01/09/2009
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Horling, Bryan C., Brukman, Michael Y., Otto, Owen, McDonnell, Philip A., Goel, Kavi J., Haahr, Paul
• Primary Examiner(s): Trujillo, James
• Assistant Examiner(s): Cheung, Hubert
• Application Number:
  US14/061,302
• Time in Patent Office: 909 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/90324 using system suggestions
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/9538 Presentation of query results

Abstract

Methods, systems, and apparatus, including computer program products, for establishing preferences for sites that are explicitly identified by a user by either a textual input or an acceptance in response to a recommendation, and adjusting the presentation of search results based on these preferences.

Claims

1. A method implemented by a data processing apparatus, the method comprising:

accessing user site data associated with a user identifier, the user site data including resource locators of resources that were selected at least once for a first user session associated with the user identifier;

identifying, from the resource locators, candidate sites for recommendation as preferred sites, each candidate site being specified by at least one of the resource locators included in the accessed user site data, the identifying comprising grouping resource locators by domain names and, for each domain name, selecting the domain name as a candidate site;

generating a recommendation score for each of the identified candidate sites based on per-site user metrics determined from the accessed user site data associated with the user identifier, wherein the per-site user metrics are associated with the user identifier for each site and are metrics based on events with respect to resources addressed by resource locators that include the domain name of the candidate site and that occurred by use of a user device during a user session associated with the user identifier, and wherein the per-site user metrics for at least one candidate site are inclusive of events with respect to at least two or more resources addressed by the resource locators that include the domain name of the candidate site;

identifying a recommended site, the recommended site being one of the identified candidate sites that has a corresponding generated recommendation score, based on the per-site user metrics associated with the user identifier, that exceeds a first recommendation threshold;

providing, in a second user session associated with the user identifier, a search results page that includes a search result that references the recommended site, and including in the search results page a recommendation to identify the recommended site as a preferred site; and

generating site preference data identifying the recommended site as a preferred site for the user identifier in response to receiving an explicit acceptance of the recommendation during the second user session associated with the user identifier, where each search result that references the domain name of the preferred site for a subsequent user session associated with the user identifier is eligible for a presentation adjustment when the search result is presented to a user device associated with the user identifier.

2. The method of claim 1, further comprising:

receiving, from a search engine, a set of search results in response to a search query received during a third user session associated with the user identifier, where each search result in the set of search results includes a resource locator that references a corresponding resource, and where each search result is ranked according to an associated score;

determining at least one search result in the set of search results includes the corresponding resource locator referencing the preferred site in the generated site preference data;

adjusting presentation of the at least one search result in the set of search results to generate an adjusted set of search results; and

providing the adjusted set of search results to the user device.

3. The method of claim 1, wherein providing the search results page further comprises including the recommendation in the search results page only if the recommendation score of the recommended site exceeds a second recommendation threshold that is greater than the first recommendation threshold.

4. The method of claim 1, wherein generating a recommendation score for each of the candidate sites comprises:

determining the per-site user metrics for each candidate site, the per-site user metrics measuring events associated with the user identifier for each candidate site;

determining per-site population metrics for each candidate site, the per-site population metrics measuring events associated with a population of user identifiers for each candidate site; and

generating the recommendation score for each of the candidate sites based on the per-site user metrics and the per-site population metrics.

5. The method of claim 4, wherein:

the per-site user metrics associated with the user identifier include one or more of;

a click-through rate measuring the rate at which search results that include a corresponding resource locator referencing the candidate site were selected in user sessions associated with the user identifier;

a long click-through rate measuring the rate at which search results that include a corresponding resource locator referencing the candidate site received a long click-through in user sessions associated with the user identifier;

an impression rate measuring how often search results that include a corresponding resource locator referencing the candidate site appear in search results are provided for a user session associated with the user identifier; and

a last-time selection value that measures a time period from a current time to a last time a resource from the candidate site was provided for a user session associated with the user identifier; and

the per-site population metrics associated with the population of user identifiers include one or more of;

a click-through rate measuring the rate at which the search results that include a corresponding resource locator referencing the candidate site were selected in user sessions associated with the population of user identifiers;

a long click-through rate measuring the rate at which search results that include a corresponding resource locator referencing the candidate site received a long click-through in user sessions associated with the population user identifiers;

an impression rate measuring how often search results that include a corresponding resource locator referencing the candidate site appear in search results are provided to the user; and

a query count measuring how many distinct queries result in the candidate site being identified in search results.

6. The method of claim 5, further comprising:

training a predictive model on the per-site user metrics and the per-site population metrics to predict a likelihood that a search result referencing the recommended site will be selected in a user session associated with the user identifier; and

wherein the first recommendation threshold is a minimum probability of selection.

7. The method of claim 1, wherein identifying, from the resource locators, candidate sites comprises identifying at least one candidate site having a resource location from which resources have never been requested for a user session associated with the user identifier.

8. The method of claim 1, wherein:

receiving from a search engine a set of search results responsive to a search query comprises receiving the search results ranked according to a first order; and

adjusting the presentation of the search result to generate an adjusted set of search results comprises adjusting the rank of the search result that includes the resource locator referencing the preferred site to rank the search results according to a second order.

9. The method of claim 8, wherein:

each search result has an associated relevance score that measures the relevance of the query to the corresponding resource specified by the corresponding resource locator; and

adjusting the rank of the search result comprises adjusting the rank of the search result only if the relevance score associated with the search result exceeds a relevance score threshold.

10. The method of claim 8, wherein adjusting the rank of the search result further comprises generating display data for the search result that causes a user device to display a visual indication that search result is an adjusted search result.

11. The method of claim 8, wherein the site preference data further identifies sites explicitly disfavored by the user, and further comprising determining that a search result includes a corresponding resource locator referencing a site explicitly disfavored and in response demoting the search result in the first order.

12. The method of claim 8, wherein:

the site preference data identifies sites by a host name or a domain name; and

adjusting the rank of the search result comprises adjusting the rank of the search result only if the site preference data identifying the site referenced by the resource locator of the search result is not a domain name.

13. A system for processing search results, comprising:

a data store storing user site data associated with a user identifier, the user site data including resource locators of resources; and

instructions stored on a non-transitory computer readable medium and executable by a data processing apparatus to cause the data processing apparatus to perform operations comprising;

accessing user site data associated with a user identifier, the user site data including resource locators of resources that were selected at least once for a first user session associated with the user identifier;

identifying, from the resource locators, candidate sites for recommendation as preferred sites, each candidate site being specified by at least one of the resource locators included in the accessed user site data, the identifying comprising grouping resource locators by domain names and, for each domain name, selecting the domain name as a candidate site;

generating a recommendation score for each of the identified candidate sites based on per-site user metrics determined from the accessed user site data associated with the user identifier, wherein the per-site user metrics are associated with the user identifier for each site and are metrics based on events with respect to resources addressed by resource locators that include the domain name of the candidate site and that occurred by use of a user device during a user session associated with the user identifier, and wherein the per-site user metrics for at least one candidate site are inclusive of events with respect to at least two or more resources addressed by the resource locators that include the domain name of the candidate site;

identifying a recommended site, the recommended site being one of the identified candidate sites that has a corresponding generated recommendation score, based on the per-site user metrics associated with the user identifier, that exceeds a first recommendation threshold;

providing, in a second user session associated with the user identifier, a search results page that includes a search result that references the recommended site, and including in the search results page a recommendation to identify the recommended site as a preferred site; and

generating site preference data identifying the recommended site as a preferred site for the user identifier in response to receiving an explicit acceptance of the recommendation during the second user session associated with the user identifier, where each search result that references the domain name of the preferred site for a subsequent user session associated with the user identifier is eligible for a presentation adjustment when the search result is presented to a user device associated with the user identifier.

14. The system of claim 13, further comprising:

receiving, from a search engine, a set of search results in response to a search query received during a third user session associated with the user identifier, where each search result in the set of search results includes a resource locator that references a corresponding resource, and where each search result is ranked according to an associated score;

determining at least one search result in the set of search results includes the corresponding resource locator referencing the preferred site in the generated site preference data;

adjusting presentation of the at least one search result in the set of search results to generate an adjusted set of search results; and

providing the adjusted set of search results to the user device.

15. The system of claim 13, wherein providing the search results page further comprises including the recommendation in the search results page only if the recommendation score of the recommended site exceeds a second recommendation threshold that is greater than the first recommendation threshold.

16. The system of claim 13, wherein generating a recommendation score for each of the candidate sites based on per-site user metrics determined from the user site data associated with the user identifier comprises:

determining the per-site user metrics for each candidate site, the per-site user metrics measuring events associated with the user identifier for each candidate site;

determining per-site population metrics for each candidate site, the per-site population metrics measuring events associated with a population of user identifiers for each candidate site; and

generating the recommendation score for each of the candidate sites based on the per-site user metrics and the per-site population metrics.

17. A computer program product, encoded on a non-transitory computer readable medium, configured to cause a data processing apparatus to perform operations comprising:

accessing user site data associated with a user identifier, the user site data including resource locators of resources that were selected at least once for a first user session associated with the user identifier;

identifying, from the resource locators, candidate sites for recommendation as preferred sites, each candidate site being specified by at least one of the resource locators included in the accessed user site data, the identifying comprising grouping resource locators by domain names and, for each domain name, selecting the domain name as a candidate site;

generating a recommendation score for each of the identified candidate sites based on per-site user metrics determined from the accessed user site data associated with the user identifier, wherein the per-site user metrics are associated with the user identifier for each site and are metrics based on events with respect to resources addressed by resource locators that include the domain name of the candidate site and that occurred by use of a user device during a user session associated with the user identifier, and wherein the per-site user metrics for at least one candidate site are inclusive of events with respect to at least two or more resources addressed by the resource locators that include the domain name of the candidate site;

identifying a recommended site, the recommended site being one of the identified candidate sites that has a corresponding generated recommendation score, based on the per-site user metrics associated with the user identifier, that exceeds a first recommendation threshold;

providing, in a second user session associated with the user identifier, a search results page that includes a search result that references the recommended site, and including in the search results page a recommendation to identify the recommended site as a preferred site; and

generating site preference data identifying the recommended site as a preferred site for the user identifier in response to receiving an explicit acceptance of the recommendation during the second user session associated with the user identifier, where each search result that references the domain name of the preferred site for a subsequent user session associated with the user identifier is eligible for a presentation adjustment when the search result is presented to a user device associated with the user identifier.

18. The computer program product of claim 17, further comprising:

receiving, from a search engine, a set of search results in response to a search query received during a third user session associated with the user identifier, where each search result in the set of search results includes a resource locator that references a corresponding resource, and where each search result is ranked according to an associated score;

determining at least one search result in the set of search results includes the corresponding resource locator referencing the preferred site in the generated site preference data;

adjusting presentation of the at least one search result in the set of search results to generate an adjusted set of search results; and

providing the adjusted set of search results to the user device.

19. The computer program product of claim 17, wherein providing the search results page further comprises including the recommendation in the search results page only if the recommendation score of the recommended site exceeds a second recommendation threshold that is greater than the first recommendation threshold.

20. The computer program product of claim 17, wherein generating a recommendation score for each of the candidate sites comprises:

determining the per-site user metrics for each candidate site, the per-site user metrics measuring events associated with the user identifier for each candidate site;

determining per-site population metrics for each candidate site, the per-site population metrics measuring events associated with a population of user identifiers for each candidate site; and

generating the recommendation score for each of the candidate sites based on the per-site user metrics and the per-site population metrics.